Key takeaways:

• The CROWN Act is legislation that protects students, workers, and housing applicants from race-based hair discrimination.
• 30 states have enacted some version of the CROWN Act, building momentum for a federal version to pass in Congress.
• Expansions of civil rights laws like the CROWN Act are more urgent than ever given Trump administration efforts to weaken nondiscrimination protections.

In an era of Trump administration attacks on nondiscrimination protections and diversity, equity, and inclusion (DEI) initiatives, workers and students need the guardrails included in the CROWN Act more than ever.

The CROWN Act stands for Creating a Respectful and Open World for Natural Hair. It expands existing civil rights law to prohibit race-based hair discrimination in schools, housing, and workplaces. Far from being solely focused on aesthetics, the CROWN Act protects against implicit bias and policing of Black and brown bodies. It codifies hair discrimination as racism and prohibits employers, public schools, and housing agencies from imposing whiteness as the uniform for success.

The CROWN Act is now law in 30 states

Pennsylvania and Rhode Island became the most recent states to pass the law in 2025, joining a growing list of states that have passed some form of the CROWN Act. Similar to the limited protections in Kentucky (which only provides public service and employee protections), Missouri also recently passed a limited version of the law that only applies to educational institutions—not workplaces or housing.

Figure A

Since the start of the second Trump administration, five states (Florida, Mississippi, Ohio, North Carolina, and South Carolina) have introduced at least one version of the CROWN Act.

Expanding CROWN Act legislation at the local level can be a significant incremental step toward expanding protections more broadly. In at least three of the states that introduced legislation, local CROWN laws have been adopted in cities including Charlotte, North Carolina, Columbus, Ohio, and Miami Beach, Florida. Such local policies have proven helpful in laying groundwork for passage of state laws elsewhere; for example, localities in Pennsylvania and Missouri passed CROWN Act laws years prior to these passing at the state level.

It’s time to pass the CROWN Act in Congress

Representative Watson Coleman (D-NJ) and Senator Cory Booker (D-NJ) introduced a federal version of the CROWN Act in the House and Senate last February. Neither bill has received a vote. With 60% of states already having passed the CROWN Act, the policy clearly has enough popular support to pass both chambers and become law.

A federal CROWN Act would provide consistent nationwide protection against hair discrimination for workers—especially Black women workers who have larger pay disparities and who are intersectionally impacted by a history of racist laws. Over 38.7% of Black women workers live in states where they are vulnerable to hair-based discrimination, based on the most recent 2024 data. That’s a slight improvement from the 44% of Black women workers who lacked protections in 2023, thanks to six more states—Kentucky, Missouri, New Hampshire, Pennsylvania, Rhode Island, and Vermont—recently passing some version of the CROWN Act.

Table 1

Although some states did not have large enough populations of Black women to be tracked in the data presented in Table 1, Black students and workers in states with smaller Black populations remain susceptible to discrimination. In South Dakota, for example, a student was told to cut their locks or leave school. Black students have shared similar stories in states across the country with varying Black population density.

Expansion of the CROWN Act must be coupled with strong enforcement

In an anti-DEI political climate, preventing hair-based discrimination must be coupled with addressing a growing crisis in federal and state nondiscrimination enforcement systems. Current threats to the Equal Employment Opportunity Commission’s (EEOC) enforcement mission are eroding protections for Black and brown workers. The EEOC was established by the Civil Rights Act of 1964 (the same law that the CROWN Act would expand) and was designed to be an independent federal agency to protect workers and promote fairness and equity in employment. The EEOC’s role in reducing discrimination has helped to boost average income by $493 to $1,233 per person since 1960. Historically, the EEOC has been a key source of protection for Black and brown people and women: A third of all EEOC charges were related to race (29,338) or sex (26,941) last year.

Figure B

Last year, the Trump-appointed chair of the EEOC hijacked the agency’s mission by actively soliciting complaints from white male workers and encouraging “DEI-related discrimination” claims of “reverse racism.” In addition to assigning a new chair, the Trump administration has weakened the EEOC by removing the ability to collect full pay data, limiting gender data collection, and removing commissioners—preventing a quorum and disrupting proceedings.

The Civil Rights Act itself and the protections that the EEOC was created to enforce are at risk. And some states are following the Trump administration’s lead by weakening state-level civil rights laws and enforcement systems. For example, in 2025, Iowa became the first state in history to strip nondiscrimination protection from a protected class when it removed gender identity from its state Civil Rights Act. In response, many cities and counties expanded their local civil rights ordinances to ensure protections against discrimination based on gender identity, but in 2026 the state went a step further and banned local governments from protecting any class not explicitly listed in state law. These recent legislative changes explicitly targeted trans people for discrimination, but will also have the broader effect of preempting local governments from adopting any expanded civil rights protections in the future, including local versions of the CROWN Act.

In the meantime, discrimination has not stopped and continues to have a measurable impact on Black, brown, and women workers—especially related to opportunities for securing a job and workplace advancement. New research finds that 59% of white adults hold anti-Black prejudice, contributing to a greater likelihood that Black applicants are rejected from jobs compared with their white counterparts. Women are paid 18.6% less than men. The disparity increases when considering race: Black women are paid 68.3% and Hispanic women are paid 64.5% of white men’s pay.

It is imperative to both expand protections like the CROWN Act at the local, state, and federal levels and strengthen state civil rights enforcement, while working toward eventual restoration of the EEOC’s capacity to carry out its mission to protect against discrimination in the workplace.

Investment in public education was one of the key reasons the U.S. became the richest country in the world in the 20th century. Universal education has many positive effects, including creating a more productive workforce and a more informed and engaged democratic society. Unions play a key role in advocating for public spending in education. Figure A shows that states with higher unionization rates spend substantially more per student on education.

Figure A

Adequate levels of per-pupil spending are instrumental policy for maintaining quality education. When per-pupil spending is low, students may struggle to get the support they need to achieve their learning goals. Districts, in turn, may have a harder time retaining teachers and staff because they can’t offer competitive salaries or benefits.

Increased per-pupil funding, by contrast, pays off for students. Greater funding for schools improves educational attainment, student achievement, and economic outcomes in adulthood. In one study, researchers found that a 10% increase in school spending for 12 years led to increases in high school graduation rates, 7% higher wages, and 10% higher family incomes in adulthood for children from districts that experienced the spending increase. More recently, researchers assessing the contribution of federal pandemic educational aid found that each $1,000 increase in per-pupil spending boosted student achievement in math and reading. Importantly, when researchers restricted their assessment to high-poverty districts, the per-dollar effect was nearly twice as large.

Unions are also instrumental to protecting quality public education. This is increasingly under threat due to state- and national-level voucher programs, which divert money away from public education and toward private schools and homeschooling. A substantial share of enrollees in voucher programs are often students who are already attending private school.

Vouchers reduce education quality in several ways. First, voucher programs yield worse academic achievement outcomes relative to public schools. Studies have found that students in voucher program schools experienced test score declines that are comparable or worse than declines due to COVID-19.

Second, voucher programs strain state budgets. And this cost comes at a time when states face budgetary pressure from the Republican tax and spending megabill (OBBBA), which reduced federal funding for Medicaid and SNAP (also known as food stamps).

Third, public school districts experience an additional hidden cost. When students leave public schools with a voucher, school districts lose revenue but must still pay the same amount for fixed costs that can’t immediately adjust to declines in enrollment, such as cooling/heating and utilities. These required payments for a district’s fixed costs mean that districts will have even less to spend on costs that can be adjusted—like school supplies or instructional support—thereby reducing services for students.

States with higher levels of unionization are more likely to have the political resources to fight back against vouchers in legislative fights. As a result, these states are less likely to have universal voucher programs, as shown in Figure B.

Figure B

In sum, there are many ways unions support workers and their communities, and their consistent support of public education shouldn’t be discounted. From advocating for enough funding for high-quality education to defending against education privatization, unions support our nation’s commitment to universal public education. This not only improves students’ academic achievement, but also their economic outcomes well into adulthood.

Paraprofessionals—who work under the supervision of licensed teachers to provide focused instructional and behavioral support to students—provide crucial services and play a critical role in educating students with disabilities. Research has demonstrated that paraprofessional educators are key contributors to effective schools.

Despite this, the current pay of paraprofessionals is too low, and clear labor shortages of paraprofessionals have developed in public schools around the country—including in New York City. Paraprofessionals working for the New York City Department of Education have annual earnings ranging from $33,000 to $54,000. EPI’s Family Budget Calculator shows that a single adult with no children in New York City needs to earn $83,262 to afford a modest but adequate standard of living. An additional $10,000 would move those salaries closer to—but still well short of—a living wage. 

One clear sign that the city could benefit from greater public investment in paraprofessionals is the fact that New York City now spends more than $1.5 billion each year on special education due process cases. These costs arise when families successfully demonstrate that the public school system has failed to provide the services required under federal law, requiring the city to cover the cost of remedies such as private school enrollment (the most well-known subset of these are often referred to as “Carter cases”).

Boosting pay by $10,000 for the roughly 25,000 full-time paraprofessionals is estimated to cost the city about $244 million in 2027, roughly 15% of what the city spends on special education due process cases. While no single policy will eliminate these due process case costs, investing in the workforce that delivers crucial special education services can certainly strengthen recruitment and retention and increase the share of families of children with disabilities who are satisfied with the quality of education and support that is being provided in public schools. In turn, more families of special education students choosing to remain in public schools will reduce expenditures on due process cases, helping defray the cost of paying paraprofessionals closer to a living wage.

Finally, we should note that while the state and city of New York would benefit greatly from increased investment in the public sector, this investment obviously requires revenue. This revenue should come from those most able to provide it: high-income households and rich corporations. The quality of services provided to the city and state will depend crucially on whether or not policymakers are willing to raise the revenue needed to make these investments.

Imagine union membership tripling in the United States. It may sound radical—if you’ve forgotten history. In fact, more than 1 in 3 private-sector workers belonged to a union in the 1950s. The results? Wages grew in tandem with the economy. The middle class thrived. Black-white wage gaps shrank. Broadly shared economic growth was a reality, not an aspiration. That’s because when workers have bargaining power, they win better wages, benefits, and working conditions. This report shines a light on what we stand to win if we rebuild union power.

Over the last four decades, big corporations and the billionaires who run them have waged a relentless campaign against unions. And they have largely succeeded in reshaping the U.S. economy. By making it harder and harder for workers to organize and bargain collectively, the rich seized more and more income and wealth, destroying the U.S. middle class. Now the wealth of the richest Americans has exploded: The richest 0.1% own more than five times the combined wealth of the entire bottom half of the country.

And yet, workers haven’t given up. In 2025, unionization ticked upward. Public approval of unions has reached some of its highest levels in decades, and more than 50 million nonunion workers say they’d join a union tomorrow if they could. That’s because they know what unions deliver. In an economy that has been rigged against working people for decades, unions serve as a counterweight to corporate power—reducing inequality and building the kind of middle class that underpins a strong and inclusive economy.

It will take serious policy change to reverse nearly 50 years of deliberate attacks on working people and their institutions. It will require that politicians stand up to the superrich and corporate interests. It will require that workers continue to build power. But, as this report shows, we have much to gain from stronger unions. An organized and empowered workforce has powerful and far-reaching economic benefits.

Nearly four decades ago, I helped found the Economic Policy Institute because working people needed a voice in the economic debates that shape their lives. This report is exactly the work we envisioned: rigorous research that puts workers at the center of economic policy, and that arms all of us with the facts to fight for them. At a time when our economy is held in relatively few hands, we need this work—and we need unions—more than ever. 

Robert Reich
Professor, writer, and former Secretary of Labor

Executive summary

Union membership in the U.S. ticked up in 2025, breaking a decades-long trend of declining unionization. But today’s unionization rate doesn’t reflect the tens of millions of workers who want a union but can’t get one. This report examines what we stand to gain if we triple current union membership to 30%—restoring it to 1950s levels, when union strength delivered rising wages, narrowing racial wage gaps, and a thriving middle class.

Tripling union membership would:

• Deliver a 14.5% raise for the median worker—amounting to more than $7,700 annually, or nearly $270,000 over a 35-year career. These life-changing increases would benefit union and nonunion workers alike.
• Shift $1.2 trillion to workers annually. This would reverse a third of the increase in inequality experienced since 1979.
• Significantly narrow racial wage gaps. Because unions tend to boost wages more for Black and Hispanic workers than for white workers, tripling union membership would close racial wage gaps by more than one-third.
• Boost the number of people with health insurance. Since unions increase other forms of compensation, like health insurance benefits, the number of nonelderly people without health insurance would fall by about 25%. Unions further reduce uninsured rates by advocating for increased public benefits like Medicaid.
• Strengthen communities. States with high union density invest more in public education, have higher unemployment insurance recipiency rates, and have all adopted Medicaid expansion.
• Protect democracy. Unions boost voter turnout, equip workers with civic skills, and actively defend voting rights. States with high union density have passed far fewer voter restriction bills than low-density states.

Download the factsheet

Roadmap for tripling union density

Reversing decades of political neglect that has stealthily undermined workers’ rights to unions and collective bargaining will require comprehensive reform that weaves together tested approaches with bold new ideas, at both federal and state levels.

Two bills with bipartisan support could help restore collective bargaining

• The Protecting the Right to Organize Act would restore private-sector workers’ right to organize and bargain collectively. It would streamline the union formation process, establish penalties for labor law violations, override so-called “right-to-work” laws, and ban “captive audience” meetings.
• The Public Service Freedom to Negotiate Act would be the first federal law that guarantees all public-sector workers at the federal, state, and local levels the right to organize and collectively bargain.

Two bold new proposals could expand the benefits of collective bargaining and help tackle the affordability crisis

• Guaranteed annual raises for newly unionized workers. Legislation providing that newly unionized workers can use arbitration to achieve a first contract (if an employer fails to negotiate in good faith) should set a minimum standard that such contracts include a cost-of-living adjustment (COLA). For the typical worker, a 3% COLA means roughly $2,000 extra a year.
• Default collective bargaining when CEO-to-worker pay ratios exceed 100:1. Declining unionization and the stratospheric rise in CEO pay are deeply connected. Strengthening the bargaining power of workers in severely imbalanced companies would enable them to capture a larger share of the wealth their work creates.

States can remove anti-union laws and protect collective bargaining

Removing so-called “right-to-work” laws and restrictions on public-sector bargaining alone would increase union density nationally from 9.9% to 14.4%. Beyond removing those unionization barriers, states can also:

• Extend collective bargaining rights to workers currently excluded from federal law (in-home child care, home health care, agricultural, and gig workers);
• Protect workers’ right to refuse mandatory, anti-union “captive audience” meetings; and
• Extend unemployment insurance eligibility to workers on strike.

Introduction

In 2025, 14.7 million workers—10% of all wage and salary workers—were union members, an increase from 9.9% in 2024.1 Though small, the increase marks a departure from prior years’ downward trend in union density and coincides with record high public favorability of unions.2

This report examines what the U.S. might look like if union membership were to triple to 30%, restoring it to roughly its 1950s peak. We analyze the economic, social, and democratic impacts of this increase; examine the potential impact on state union density rates if all states were to remove anti-union policies; and offer policy recommendations to expand union membership.

While tripling union membership is an ambitious goal, it is fully consistent with workers’ own demand for unions. Recent survey data show that 43% of nonunion workers would vote to unionize if given the opportunity—the equivalent of about 56 million wage and salary workers (Ahlquist, Grumbach, and Kochan 2024; McNicholas, Poydock, and Shierholz 2026). If all these workers unionized, union density would rise from 10% to 48.7%—well above the 30% goal we examine in this paper.

Union decline, wage suppression, and affordability

As union membership has declined, workers’ wages have been suppressed and inequality has skyrocketed. The gap between typical workers’ pay and economy-wide productivity is at a historic high.3 Since 1979, productivity has grown 2.7 times as much as pay for typical workers (Figure A). This divergence reflects both rising wage inequality—high earners experiencing much stronger wage growth than typical workers—and a shrinking share of the economy’s income going to workers overall. Between 1979 and 2023, real (inflation-adjusted) earnings for the top 0.1% grew 354% (from $618,000 to $2.8 million), while earnings for the bottom 90% of households grew just 44% (from $30,000 to $43,000) (EPI 2026a). And workers are now taking home a historically low share of corporate-sector income—meaning shareholders and other capital owners are capturing more than ever before (EPI 2026b).

Figure A

Inequality has grown since 1979 because of intentional policy choices that suppressed wages for typical workers to accelerate income growth at the top. If pay for typical workers had kept pace with productivity over the past nearly five decades, their paychecks today would be more than 40% larger (EPI 2026c). Though affordability pressures are often framed as a problem of high prices, the real problem is this wage shortfall—and reversing that shortfall must be central to any serious affordability agenda. Collective bargaining is the most effective mechanism workers have to raise their wages and secure their fair share of economic growth. Expanding union membership and collective bargaining is central to addressing the affordability squeeze.

Unions raise wages for all workers

One of the most well-studied benefits of unions is the ability of collective bargaining to raise pay—resulting in a “union wage premium.”4 Historically, that premium has fluctuated around 15–20% (EPI 2026d). Unions also reduce inequality by compressing the wage distribution within firms, with the largest benefits going to workers at the bottom and middle. This is one reason the union wage premium is typically larger for those with less education, and for Black and Hispanic workers relative to white workers (Bivens et al. 2023).

Less appreciated is that unions also boost wages for nonunion workers. Higher wages at unionized firms make those jobs more attractive to nonunion workers, indirectly pressuring nonunion firms to raise wages to keep their workers. Nonunion firms may also raise wages to avoid unionization, especially when union density is already high in the relevant sectors and occupations.

Finally, unions also advocate for progressive changes in the tax and benefit system—indirectly raising pre-tax pay for low- and middle-wage workers. This happens in two main ways: First, when taxes on top incomes are higher, corporate executives have less incentive to maximize their own income at the expense of workers’ wages.5 Second, more generous social benefits like SNAP and Medicaid increase the bargaining power of lower-paid workers by making them less economically vulnerable and better able to reject low-quality jobs (Bivens 2026).6

These “spillover” effects of unions on the nonunion sector are significant and increase when union density is high. Rosenfeld, Denice, and Laird (2016) and Farber et al. (2021) find strong evidence that unions raise the wages of nonunion workers. Consistent with this, Fortin, Lemieux, and Lloyd (2021) estimate that between 1979 and 2017, declining unionization was responsible for 37% of the increase in inequality between the 90th and 50th percentiles of male workers—and more than half of that impact occurred because, as unions weakened, nonunion employers faced less pressure to offer higher wages.

Tripling union density would shift $1.2 trillion a year to working people

Our analysis of the relationship between union density and state median wages finds that for every 10 percentage point difference in union density, real median wages are about 7.2% higher.7 If we tripled union density from 10% to 30%, the real median wage for all workers (including both union and nonunion workers) would rise from $25.67 in 2025 to $29.39 per hour—a 14.5% increase.8 A full-time, full-year worker earning that wage would see annual pay rise by more than $7,700.

To put this into perspective, consider the substantial growth in the gap between pay and productivity between 1979 and 2025. A full-time, full-year worker at the 2025 median wage was paid about $53,400 per year—but would have earned about $76,400 had their pay kept pace with productivity growth (EPI 2026e). Through direct wage increases for union members and unionization’s spillover effects on nonunion workers, tripling union density would close roughly one-third of the productivity-pay gap, increasing annual pay to $61,100.

To contextualize the gain in another way, consider that the median annual cost of a mortgage in the U.S. is $18,252. An additional $7,700 per year would cover more than 40% of that cost.9

Perhaps more strikingly, consider that an extra $7,700 a year adds up to nearly $270,000 (in today’s dollars) over the course of a 35-year career—a life-changing increase for a working family. Adjusting the USDA’s latest estimate of the cost of raising a child to 2025 dollars yields a current cost of roughly $317,000 to raise a child from birth through age 17, including food, housing, transportation, health care, clothing, child care, education, and other expenses (USDA 2017). The additional $270,000 in career earnings that a typical worker would receive if union density were tripled would cover 85% of that cost.

Or, according to the College Board, the average annual cost of attending a four-year in-state public college, including room and board, is about $31,000 (Ma, Pender, and Hu 2025). Over four years, that comes to $124,000 per student—so the additional $270,000 in career earnings would more than cover the cost of sending two children to college.

Yet another way to show how transformational these effects would be for the U.S. economy and the economic security of typical families: Assuming the 14.5% wage boost we estimate from a tripling of union density applies to the bottom 80% of the U.S. workforce, this would raise these workers’ combined pay each year by $1.2 trillion.10 

Several useful reference points help illustrate the scale of this $1.2 trillion increase in workers’ annual pay. First, the increase would be nearly three times as large as the $430 billion U.S. families received in “COVID checks”—and unlike those one-time checks, these gains would be permanent and recurring.11 Second, the $1.2 trillion that tripling union density would shift to workers slightly exceeds the Pentagon’s 2025 budget—a redistribution of income on a scale comparable to the entirety of the U.S. military-industrial complex.12 Third and most revealingly, this increase would claw back a full third of the increase in income inequality since 1979. Between 1979 and 2022, the share of market income going to the richest 20% of households rose by 12.0 percentage points (with nearly all of that increase—10.1 percentage points—accruing to the richest 1%). A $1.2 trillion increase in the incomes of the bottom 80% of households would reverse a third of that shift, 4.0 of the 12.0 percentage points.13

Our estimates likely underestimate future gains

It’s worth noting that while our estimates show large wage gains to workers from tripling union density, they are likely being significantly held down by the fact that union density in the U.S. is starting from such a low level (10%). Our estimates therefore likely understate the wage gains workers will experience once the labor market reaches a higher level of union density. Fortin, Lemieux, and Lloyd (2021) find that the wage payoff to union density is much larger where density is already high, and this shows up clearly in our analysis as well: Below 15% union density, a one percentage point increase in density is associated with just a 0.2% increase in the median wage; above 15%, the median wage increases by 0.9%—a wage response more than four times as large (see Figure B).14

The economic intuition behind this 15% threshold is that a minimum scale of unionization is needed for collectively bargained wages to exert a strong influence on pay scales—both union and nonunion—throughout the broader economy. While every additional percentage point of unionization gives another 1% of workers the union pay premium, it has a more limited effect on wage-setting throughout the broader labor market until overall union density reaches around 15%.

Today’s typical state has a union density of about 10%, which means the wage gains from each additional percentage point of unionization—while still significant—are muted relative to the gains that occur once union density exceeds 15%. In essence, the policy assault on unions in recent decades has pushed density so low that we need a 50% increase—from 10% to 15% union density—just to start realizing the “normal” gains unions can bring to a high-road economy. Raising density from 10% to 30% would carry the labor market across that 15% threshold and well into the higher-density regime, where each additional percentage point increase in unionization delivers substantially larger gains for the median worker. This is why a transformational policy effort is needed: We must rebuild union density from the depressed levels created by decades of political neglect to the point where the full wage benefits of widespread collective bargaining can be realized.

Unions reduce racial wage gaps

Because unions increase wages for Black and Hispanic workers more than for white workers, unions reduce racial wage gaps. And this dynamic is not new—research shows that the union premium was already higher for Black workers than for white workers by the mid-20th century, meaning that the spread of unionization in that era was one of the country’s greatest forces for racial justice (Farber et al. 2021; Bivens et al. 2023). But with the decline in unionization over the past 45 years, racial wage gaps have widened and are now substantially worse than they were in 1979—damage that restoring union power would more than reverse. In 2025, the median Black or Hispanic worker earned 76.5% of the median white worker wage—a wage gap of 23.5%. Tripling union density would close more than one-third of that gap, reducing it to 14.6%.15

Unions boost benefits, not just pay

Unions also increase other forms of compensation, like health insurance benefits. In 2025, 95% of union workers had employer-sponsored health insurance, relative to 71% of nonunion workers. In addition, as discussed later, unions further reduce uninsured rates by advocating for increased public benefits like Medicaid. Figure C shows that nonelderly uninsurance rates are substantially lower in states with higher union density. To do this analysis, we divided the 50 U.S. states plus the District of Columbia into three equally sized groups based on their current (2023–2025) level of union density.16 Most low-union-density states have double-digit uninsured rates, but these rates typically fall to the single digits in medium- and high-union-density states. If union density tripled and all states were high-density states, the national nonelderly uninsured rate would fall by about a quarter, from 9.8% to 7.3%.

Figure C

Unions benefit communities

Unions don’t just improve workers’ pay and benefits—they give workers a voice in shaping the social fabric of their communities, lifting standards for union and nonunion workers alike. The data reveal a strong correlation between higher levels of unionization and a range of economic and personal well-being measures. When more workers are in unions, states have more equitable economic and social structures (McNicholas et al. 2025). Meanwhile, states with anti-union policies have lower union density and consistently worse outcomes across these measures.

Economic well-being

We first examine economic well-being—broadly defined as working people having the means to support themselves—using two indicators: the gap between productivity and pay, and unemployment insurance recipiency rates.

As discussed earlier, there has been a growing gap since 1979 between productivity—the value of the output generated by a worker in the U.S. economy in an hour of work, on average—and typical workers’ hourly compensation. Far from accidental, this divergence stems from deliberate policy choices in the U.S. beginning around 1980 that allowed the gains from economic growth to be increasingly captured by the highest earners, shareholders, and other capital owners—rather than being broadly shared across low-, middle-, and high-income households (Shierholz 2024).

The degree to which productivity gains have translated into higher pay for typical workers varies widely across states. Figure D shows that in states with smaller declines in unionization since 1979, a larger share of productivity gains translated into higher pay for typical workers. In other words, in places where unions remained stronger, it wasn’t just those with high incomes and wealth who benefited from economic growth—working people, both unionized and nonunionized, saw a bigger share of the gains.

Figure D

Unemployment insurance (UI) recipiency rates also offer insight into economic well-being. When a worker is laid off and their household income falls, unemployment insurance supports them and their family until they find another job. UI is a joint federal-state program that relies on state UI systems to effectively deliver benefits to unemployed workers. States differ in their rules and approaches under the federal UI framework, leading to wide variation in the share of unemployed workers receiving UI benefits (the UI recipiency rate).

Studies show a strong correlation between higher UI recipiency rates and high union density (Clegg et al. 2022); Figure E illustrates this strong positive relationship. In recent years, the average UI recipiency rate in high-union-density states was 36%, double the 18% rate in low-union-density states.

Unions fight to increase UI eligibility, ease of access, benefit levels, and benefit duration—to the benefit of all workers, union and nonunion (Hertel-Fernandez and Gould-Werth 2020). The strong correlation between union density and state UI recipiency rates suggests that expanding unionization by removing anti-union policies would result in higher UI recipiency rates.

Figure E

Personal health and well-being

Next we examine personal health and well-being—which we define as an individual’s physical and mental health—using one indicator, Medicaid expansion.

As we noted earlier, almost all union workers (95%) have access to health insurance, and uninsured rates are lower in high-union-density states. To show how unions influence health care coverage beyond explicitly bargained workplace plans, we examine whether a state has expanded Medicaid under the Affordable Care Act to broaden eligibility. The Medicaid public insurance program provides critical health coverage to families and individuals with low incomes. Research has found that Medicaid expansion has expanded access to health care; improved health outcomes, including fewer premature deaths; lowered uncompensated costs; bolstered financial security; and boosted economic mobility (CBPP 2020). While the Medicaid program has broad federal guidelines, it is administered by states—and a 2012 Supreme Court decision gave them the power to decide on eligibility and expansion. Since 2014, 40 states plus the District of Columbia have adopted Medicaid expansion; 10 states have not (Alabama, Florida, Georgia, Kansas, Mississippi, South Carolina, Tennessee, Texas, Wisconsin, and Wyoming).

Figure F shows a stark divide between high- and low-union-density states in terms of Medicaid expansion. All 17 high-union-density states have expanded Medicaid. Out of 17 medium-union-density states, 14 have adopted Medicaid expansion (Alabama, Kansas, and Wisconsin are the only medium-union-density states that have not adopted expansion). Just 10 out of 17 low-union-density states have adopted Medicaid expansion (Florida, Georgia, Mississippi, South Carolina, Tennessee, Texas, and Wyoming are the low-union-density states that have not adopted expansion).

There is a strong correlation between union density and Medicaid expansion. If states removed anti-union policies and became high-union-density states, they would be joining a group of states in which Medicaid expansion is currently universal, potentially helping to close the Medicaid coverage gap that currently leaves 1.5 million adults uninsured (Zhang 2026).

Figure F

Spending on public education

Union density also intersects with public education investment. Investment in public education helped make the U.S. the richest country in the world in the 20th century. The benefits of universal education are many—including a more productive workforce, and a more informed and engaged democratic society equipped to think critically. Unions play a key role in negotiating public spending on education. At the school district and state level, education unions advocate for increased funds for schools. 

We can measure the relationship between union density and public education investment by looking at states’ level of per-pupil spending. Adequate levels of per-pupil spending are crucial to maintaining quality education; low per-pupil spending limits the support students receive toward achieving their learning goals and makes it harder for districts to retain teachers and staff by offering competitive salaries or benefits. Increased per-pupil funding pays off for students. Research has consistently shown that increased funding for schools improves educational attainment, student achievement, and economic outcomes in adulthood (Jackson and Mackevicius 2023). In one study, researchers examined the impact of school finance reforms between 1972 and 2010 and found that a 10% increase in school spending for 12 years leads to increased high school graduation rates, 7% higher wages, and 10% higher family incomes in adulthood for children from districts that saw the spending increase (Jackson, Johnson, and Persico 2016).

Figure G shows that states with higher rates of unionization spend substantially more per pupil on education—demonstrating the impact of union advocacy on education investment and building a more educated society.

Figure G

Voting rights and democratic participation

Many states disproportionately disenfranchise people of color (Johnson and Feldman 2020), and the number of voter suppression laws enacted across the country is high and rising. These restrictions threaten democratic participation and the stability of our democracy.

Unions strengthen democracy by encouraging civic engagement and helping workers become informed voters. Unions communicate with their members about issues and candidates to make sure workers have the information they need when they go to the polls. And as democratic institutions where members elect leaders and ratify contracts, unions serve as “schools of democracy,” equipping workers with civic skills that carry over to public life (Andrias and Hertel-Fernandez 2026). More broadly, because many organizations that shape policy represent elite or corporate interests, unions are a countervailing force, bringing working people’s voices into legislative debates and building coalitions that can win against well-resourced opposition.

Research shows that these effects translate into greater political participation. Union members are more likely to vote than the general public, and voter turnout is higher in states with greater unionization (EPI 2021). Conversely, turnout is lower in states with “right-to-work” (RTW) laws, which weaken unions and in turn the civic engagement they foster. Feigenbaum, Hertel-Fernandez, and Williamson (2018) found that RTW laws reduced presidential election turnout by 2%, a substantial effect given the narrow margins that often decide presidential elections.

We build on this evidence by examining the relationship between union strength and laws that restrict voting. Figure H shows that there is a strong correlation between union density and voter suppression legislation. A large majority of low- and medium-union-density states passed at least one voter restriction bill between 2023 and 2025, while a large majority of high-union-density states passed none. Among high-union-density states, 14 out of 17 did not pass any voter restrictions during this period, while only eight of the medium-union-density states and three of the low-union-density states can claim this distinction.

These results are consistent with evidence that unions actively defend voting rights. Unions have long mobilized members to oppose voter suppression laws, educated communities about their rights, and turned out to vote despite increasing barriers (Bondy 2025). Unions have also expanded access to voting: Dean, McCallum, and Grumbach (2023) found that county-level union density was associated with greater access to ballot drop boxes—a highly secure way to increase access to voting—during the 2022 midterm elections. Together, these findings suggest that stronger unions not only increase voter participation but also help preserve and expand access to the ballot.

Figure H

State policy and union density

Increasing union density across all states—including those where anti-union policies have historically suppressed unionization—is essential to strengthening both the economy and democracy, improving all workers’ wages, and expanding access to public goods like health care, retirement, and education.

As the previous section made clear, residents of high- and low-union-density states face starkly different economic and social realities. These wide variations in union density across states—and highly variable economic outcomes—are strongly correlated with variations in state labor policies. For example, the union rights of millions of public-sector, agricultural, and domestic workers (including home health care and in-home child care providers) vary by state because workers in these occupations are excluded from coverage under federal law. And even for private-sector workers whose union rights are covered under federal law, collective bargaining rights are limited by anti-union “right-to-work” laws in many states. Research shows that anti-union state policies result in lower union density. Union density averages 6.2% in states with anti-union RTW laws compared with 15.8% in non-RTW states, and states that limit or prohibit public-sector collective bargaining have lower overall union density than other states (Sherer and Gould 2024; Sherer and Morrissey 2026).

Here, we consider what union density would be if states with certain anti-union laws adopted the policies of states that better protect workers’ rights to unionize and collectively bargain. We consider two policy changes: 1) requiring state and local governments to bargain with public-sector workers who choose to unionize; and 2) repealing so-called “right-to-work” laws that limit collective bargaining rights of private-sector workers. These are two of many policies that states can adopt to remove barriers to unionization, but we focus on them because of their clearly documented, large-scale impact on large numbers of workers across much of the country. At present, roughly half of U.S. states have RTW laws in effect and do not require public employers to bargain with unionized public employees. Modeling changes to these two policies lets us estimate what would happen to union density if we removed large state-by-state labor policy disparities and created a stable “floor” for collective bargaining for workers across all states. This provides an important starting point for considering how additional, stronger state labor policies might affect union density, worker wages, and democracy.

Public-sector workers face wide disparities in bargaining rights across states

Variations in collective bargaining rights for public-sector workers are rooted in history, notably in Southern states where systemic racism drove the enactment of anti-union policies (Childers 2023; Dixon 2007; Kaufman 2018; Stelzner, Hoyt, and Ramchurn 2019). By the 1980s, most states had enacted public-sector collective bargaining policies, though these statutes varied in strength and coverage (Freeman and Ichniowski 1988). In more recent decades, many states have repealed or harshly limited formerly robust public-sector collective bargaining statutes, while a few states have expanded them. A key distinction is whether state and local government employers have a legal duty to bargain with unionized employees over pay and working conditions, rather than being merely permitted to bargain—or outright banned from doing so.

Sherer and Morrissey (2026) classify state public-sector bargaining rules across five categories of workers: teachers, police officers, firefighters, other local government employees, and other state government employees. States vary widely in which categories they cover and how strongly, and that variation maps closely onto union density. In states with strong, uniform duty-to-bargain rules across all five categories, public-sector union density can be close to 50% or higher, while in states that ban bargaining across most categories, state and local density can fall as low as the single digits. Pay consequences track these differences. Across the country, public-sector employees earn less than their private-sector counterparts with similar experience and education. This public-sector pay gap is smaller when bargaining rights are strong: Sherer and Morrissey (2026) estimate that in states with robust duty-to-bargain rights, public-sector workers earn 14.3% less than private-sector workers; the gap widens to 19.6% in states where bargaining is only permitted, and to 22.5% in states where it is banned.

“Right-to-work” laws weaken union density and suppress wages for all workers

“Right-to-work” laws prohibit unions and employers from negotiating over union security—the contract terms under which workers covered by a union contract either join the union or pay an agency fee covering their share of representation costs. State RTW laws first emerged as part of anti-union industry campaigns to suppress multiracial worker organizing and maintain Jim Crow labor relations in Southern states following the passage of the National Labor Relations Act (Pierce 2017). By making union finances more unpredictable and leaving unions more vulnerable to ongoing employer interference, these laws weaken unions’ stability, suppress union density, and limit workers’ bargaining power. As a result, RTW laws generate negative outcomes for all workers—with or without a union. States with RTW laws have lower unionization rates and higher income inequality, and workers in RTW states on average have lower wages and benefits (Fortin, Lemieux, and Lloyd 2022). Sherer and Gould (2024) document that workers in RTW states earn 3.2% less than otherwise comparable workers in non-RTW states—about $1,670 per year for a full-time worker.

There are currently 27 states with RTW laws in place—including Colorado, which is not a traditional RTW state, but whose law has, since 1943, effectively imposed RTW conditions. By barring unionized workers from negotiating over union security unless they first win a second, state-administered election by a 75% supermajority of those voting or 50% plus one of all eligible workers—whichever is higher—Colorado’s anti-union policy has produced RTW-like outcomes, so we group it with RTW states.

Removing state anti-union policies would raise national union density from 9.9% to 14.4%

We estimate that these two policy changes—1) requiring state and local governments to bargain with public-sector workers who choose to unionize, and 2) repealing RTW—would substantially increase unionization in states that currently have more anti-union policies in place, as shown in Table 1.17 In the 24 states without a duty to bargain with public-sector employees, establishing that duty would raise public-sector union density from 16.7% to 46.7%. In the 27 RTW states, repealing RTW would increase private-sector density from 3.4% to 8.8%. Overall, removing these anti-union policies would increase union density in 27 states from 5.4% to 14.3%. Nationally, union density would rise from 9.9% to 14.4%.18 This is a conservative national estimate reflecting changes in union density only for the 27 states undergoing hypothetical policy changes and not including likely impacts on the 23 states and D.C. where stronger union policies are already in place. These states would likely also see some level of increased union density under this scenario, given that removal of anti-union state policies across the country would strengthen worker bargaining power and union organizing resources in all states.

Table 1

Removing state-level anti-union policies would increase median wages by up to $4,900 a year

Were the 27 RTW states to increase unionization to estimated levels, overall annualized median wages would rise by $1,600 to $4,900 for full-time, full-year workers, depending on the state. Table 1 shows that if Virginia’s density rose from 4.9% to 15.5% because of the removal of state-level anti-union policies, annual pay for full-time, full-year workers at the state median wage would increase by about $4,500. Removing these anti-union policies in Texas could cause a similar increase in density, raising annual pay by $3,400.

Of course, were workers in all states able to achieve 30% union density, the effects on median pay would be even larger. These impacts are shown in Figure I. The 23 percentage point increase in density in Alabama, for example, would increase annualized median pay there for full-time, full-year workers by about $8,500.

Figure I

Recent state policy changes show that anti-union laws lower union density

Recent “natural experiments” in states that have changed one or both of these labor policies also show the impact on union density. For example, in states that adopted new RTW laws since 2010, union density fell 3.8 percentage points between 2010 and 2023. By comparison, union density declined only 2 percentage points during this same period in non-RTW states (Sherer and Gould 2024).

Wisconsin is an example of a state that both repealed a previous duty to bargain with public employees (in 2011) and enacted a new RTW law (in 2015). Since adopting these anti-union policies, between 2011 and 2024, Wisconsin’s union density has fallen by half, from 14% to 7%. Over this same period, national union density declined just 2 percentage points, from 13% to 11%. With both a RTW law in place and no public-sector duty to bargain, Wisconsin’s union density is plummeting in comparison with national and regional trends (Dresser, Rogers, and Vasquez 2025). Other studies show that Wisconsin’s adoption of anti-labor policies and declining union density have generated a wide array of harmful outcomes for the state’s economy and democracy, including sharp decreases in public education spending, declining worker wages, widening public-sector pay gaps (including stark increases in gender pay gaps for teachers), and declining voter participation (Nack et al. 2019; García and Han 2021; Biasi and Sarsons 2022; Feigenbaum, Hertel-Fernandez, and Williamson 2018).

States have other policy options to enable workers to increase union density

Beyond these two policy changes, states have numerous other opportunities to increase union density by removing obstacles to unionization and ensuring more workers have full rights to collectively bargain (Sherer 2026). A few examples that have not yet been adopted in most states include:

• Creating pathways to collective bargaining for all workers in occupations not currently covered by federal labor law, including in-home child care and home health care workers, agricultural workers, and rideshare or delivery drivers treated as “independent contractors” by digital platform companies;
• Protecting workers’ freedom to refuse mandatory “captive audience” meetings on political or religious topics not related to work duties, including mandatory anti-union meetings employers typically hold when attempting to block workers from unionizing;
• Ensuring workers whose paychecks stop due to a strike or lockout are eligible to apply for unemployment insurance; and
• Establishing or expanding state agency capacities and public labor education programs that ensure workers can easily learn about their union rights.

We do not model the specific impacts of all possible state policy changes in this report. But our conservative, baseline estimate of increases in union density likely to result from removing two of the largest state obstacles to workers’ union rights—RTW laws in 27 states and the lack of a duty to bargain for public employees in 24 states—suggests that combining these changes with additional policy changes would likely lead to additional increases in union density across the country.

Federal policy and union density

Given the clear benefits of tripling union membership, it is critical that policymakers prioritize reforms that enable workers to organize and collectively bargain. It is tempting to search for a policy “silver bullet” to restore the promise of the National Labor Relations Act (NLRA), our primary labor law. But the reality is that policymakers must adopt comprehensive labor law reforms to eradicate the effects of decades of political neglect that has steadily undermined the effectiveness of U.S. labor law. As discussed above, these reforms must at a minimum reverse state anti-union policies, via state or ideally federal labor law reforms that end Jim Crow-era occupational exclusions, ensure equal union rights and pathways to collective bargaining for all workers in all states, and eliminate RTW laws.

Labor law reform has passed the House three times with bipartisan support

A consistent narrative maintains that such policy change is impossible, but that narrative relies on the mistaken notions that real reforms require more political power than the labor movement has and that our legislative process prevents labor law reform. It is true that winning policy fights takes significant political capital. But labor law reform must not be seen as solely a labor movement priority. It is central to any attempt to address affordability pressures and economic inequality. And in this moment, such reform is imperative to our democracy. While the Senate filibuster rules do prevent much legislative progress, they are not inevitable dictates but rather agreed-upon rules that, as history shows, can change. It is deeply destructive to our political system to convince working people that the system simply cannot be expected to serve their interests, which is what this flawed narrative around labor law reform reinforces. To be clear, policies favored by elites benefiting from the unequal status quo are far more likely to become law than even broadly popular ones. But these reforms have the potential to help rebalance this inequality.

In fact, in the last two decades, legislation to significantly reform labor law has passed the U.S. House of Representatives three times—with bipartisan support each time. Even in the current Congress, legislation aimed at restoring collective bargaining rights has managed to bypass Republican House leadership opposition via parliamentary procedure (discharge petition) and pass with bipartisan support.19 Of the on average 12,000 bills introduced in a given Congress, labor law reform measures have had more success than roughly 95% of bills—passing a chamber with bipartisan support. Of course, Senate rules around the filibuster requiring 60 votes to proceed on consideration of various measures have been a significant impediment to the Senate passing these reforms. However, since 2013, the Senate has used the “constitutional option” to change its own rules for procedures and has eliminated the 60-vote requirement to proceed on specific debates three times. That could be done here, too.

Record high union approval strengthens the case for labor law reform

Labor unions today have significantly higher public approval ratings than when major labor law reform was last considered, polling more than 20 points higher.20 Since 2021, approval for unions has remained high, with over 68% of people in the U.S. viewing unions favorably (Brenan 2025). This positive view of unions is shared across generations, with majorities of Boomers (59%), Gen X (58%), Millennials (61%), and Gen Z (63%) viewing unions favorably. Young adults (ages 18–35) have the highest favorability rate at 72% (Glass 2025).

Unions are viewed positively across party lines, with both Democrats (90%) and independents (69%) having high favorability rates for unions, and over 40% of Republicans approving of labor unions (Brenan 2025). Some conservative organizations recognize that unions are popular among workers: Research by American Compass (2025) finds that at least 46% of Republicans view unions somewhat favorably, with favorability increasing among young Republicans (60%).

Data from the American National Election Studies show that people in the U.S. favor unions over big business now more than ever—with the average rating for labor unions hitting a new high (60%), while big business hit a low (44%) (Sojourner and Reich 2025). Further, most people in the U.S. say the decline in union density is bad for the country (60%) and bad for working people (62%). Most young adults (69%), including young Republicans (52%) and young Democrats (82%), view the decline in union density as negative for working people (Van Green 2025).

Three bipartisan bills would pave the way toward restoring collective bargaining rights

As with many popular policies, high public opinion of unions has not translated into the enactment of labor law reform. Even though several bills that would reform aspects of our current labor law are being introduced and even passing with bipartisan support in the U.S. House of Representatives, the Senate has failed to act on the legislation. There are three main bills with bipartisan support that would help reform our current labor law system and each should be passed.21

• The Protecting the Right to Organize (PRO) Act would restore private-sector workers’ right to organize and bargain collectively. It would streamline the process of forming a union, create a roadmap to reaching a first collective bargaining agreement, establish civil monetary penalties for employers who violate labor law, among other reforms—such as expanding coverage to more workers, overriding state “right-to-work” laws, and banning “captive audience” meetings (McNicholas, Poydock, and Rhinehart 2021). Since its first introduction in 2019, the PRO Act has passed the House of Representatives twice with bipartisan support.
• The Public Service Freedom to Negotiate Act (PSFNA) would establish public-sector workers’ right to organize and bargain collectively. This would fill an important gap—there is currently no federal law that guarantees all public-sector workers at the federal, state, and local levels the right to organize and collectively bargain. As a result, more than half of U.S. states today lack comprehensive collective bargaining laws for state and local public-sector workers (Sherer 2026).
• The Protecting America’s Workforce Act (PAWA) would reverse President Trump’s executive order that revokes collective bargaining rights for federal workers. The bill would reestablish collective bargaining rights for more than 1 million federal workers. In December 2025, PAWA passed the House of Representatives by a 231-195 vote.

Beyond passing these important bills, policymakers should also consider additional policies to help workers access collective bargaining and union membership. We propose two concrete policies that complement the labor reforms above. While these two proposals alone will not result in tripling union membership, they could help more workers receive the benefits of collective bargaining, which serves as a corrective to the affordability crunch and the economic inequality that characterizes our economy.

Guaranteed annual raises for workers in first contract arbitration

When workers win a union and begin to collectively bargain with their employer, both parties have an obligation to bargain in “good faith.” However, given that the NLRA lacks meaningful penalties for violations, employers often engage in bad faith bargaining tactics with few consequences. In fact, it is well documented that many union-busting consultants view bargaining as a new phase of union-busting and advise employers to drag out bargaining for as long as possible. The goal is to avoid reaching a contract for a year or more, hoping that workers will then give up their union through decertification, which is permitted one year after the election. It is no coincidence that workers who have successfully formed a new union spend an average of 465 days bargaining for a first contract before reaching an agreement.

The Faster Labor Contracts Act, which overwhelmingly passed the U.S. House of Representatives last month, would discourage delay and promote good faith bargaining by establishing a mediation and binding arbitration process when employers refuse to bargain in good faith. Beyond guaranteeing workers first contract arbitration, we suggest the adoption of a mandatory cost-of-living adjustment (COLA) for workers whose first contract bargaining ends up in arbitration. This would establish an important guardrail for workers in the arbitration process, given that an arbitrator will ultimately impose a final contract on both parties. This provision would ensure that workers receive at least a COLA each year of their first contract. In other words, by voting for a union, workers would be voting for a guaranteed raise.22

For the average production, nonsupervisory worker, a 3% COLA would result in almost a $2,000 increase in annual earnings for a full-time, full-year worker. 23 Table 2 shows that, while these increases would vary widely by industry because pay varies widely by industry, workers in some of the lowest-paid industries would still see annual pay increases above $1,000.

Table 2

Default collective bargaining when CEO-to-worker pay ratios exceed 100:1

The outsized and ballooning gap in recent decades between CEO pay and the pay of typical workers is among the clearest signs that the economy’s rules have been rigged to allow the most powerful to claim oversized shares of income growth. This CEO-to-worker pay ratio rose nearly tenfold from 1978 to 2024, as the pay of CEOs rose over 1,000% while the pay of typical workers rose just 26%. The explosion of CEO pay was driven not by CEO’s rising productivity or skills, but by policy changes that boosted both the incentives and the ability of CEOs to raise their own pay.

When a firm’s CEO makes more than 100 times what the typical worker in their industry makes, effective corporate governance has clearly broken down—which threatens the ability of those outside the C-suite to receive their fair share of the firm’s income. The two main parties competing with executives to claim a share of a firm’s income are the shareholders and the firm’s workers. The best mechanism for ensuring that workers are able to bargain against CEOs and shareholders for their fair share of the income generated by the firm is collective bargaining.

We propose instituting default collective bargaining in any firm that meets or exceeds this 100:1 ratio. Securities and Exchange Commission (SEC) reporting already requires the annual disclosure of CEO pay (SEC 2007). We propose that these disclosures be reported not only to the SEC but also to the National Labor Relations Board. Comparing those disclosures with Bureau of Labor Statistics data on the pay of a typical worker economy-wide in various industries, the NLRB would determine which corporations meet the default collective bargaining ratio of 100:1.24 The NLRB would then publish a public notice and notify specific corporations and their workers that the default collective bargaining ratio had been met, triggering default collective bargaining at the firm. The NLRB would then direct an election for a bargaining representative to be held within 45 days of the notification.25

Nonsupervisory workers at these firms would then have the opportunity to select a bargaining representative of their choosing—whether a labor organization or other representative. The NLRB would conduct the election for and certify a bargaining representative, after which bargaining would start. For purposes of default collective bargaining, the presumption for the bargaining unit would be a wall-to-wall unit—a model where all nonsupervisory employees join a single bargaining unit—ensuring comprehensive coverage under the collective bargaining process and resulting contract. This standard would apply absent an employer or employee representative demonstrating extraordinary circumstances.26 The NLRB would then determine the appropriate bargaining unit, the same way the agency already accepts or rejects bargaining unit proposals. We propose that all such questions be resolved within 10 days of the NLRB’s notice of election for bargaining representative. This proposal would not apply to existing bargaining units, but in firms where unions are present, it would apply to any nonunionized nonsupervisory employees, as described below.

A threshold of 100:1 for the CEO-to-typical-worker pay ratio is reasonable and fair. A ratio this high clearly signals a fatal breakdown in the ability of the firm’s workers and shareholders to protect their income claims against executives looking to maximize their own incomes. A ratio this high will not affect most publicly traded companies but will be relevant to a large share of the U.S. workforce.

Comprehensive data on CEO pay are available only for a subset of firms: The Compustat Execucomp database tracks CEO pay for half of publicly owned companies (and most U.S. firms are not publicly owned). For the firms in the Compustat database, under half (41.7%) had a CEO-to-typical-worker pay ratio exceeding 100:1 in 2024. This proposal would provide the 30 million workers employed in firms that have higher ratios—representing about a fifth of total private-sector employment in the United States—with collective bargaining.

It is not a coincidence that the stratospheric rise in CEO pay was accompanied by an acceleration in the downward trend of unionization in the United States. The broader correlation between declining unionization and the rise of inequality is well established, and the highest-quality research documents that this relationship is clearly causal, with the decline of unions leading to a higher share of income claimed by households at the top of the income distribution (Farber et al. 2021).

To be clear, using a threshold CEO-to-worker pay ratio as a trigger for default collective bargaining will not alone solve the problem of excess CEO pay. We focus on this threshold as a policy trigger because it is a readily available and measurable indicator of the outsized control of executives over the distribution of the firm’s income. In the total absence of collective bargaining or any other strong corporate governance institutions, firm executives will shift income in a zero-sum fashion away from both the workers and the shareholders of a firm. If shareholders manage to assert some influence over CEO pay—e.g., through public pressure campaigns or say-on-pay mandates—one likely outcome will be CEOs looking to suppress the firm’s wages even further to make room for both their own outsized salaries and a return to shareholders that placates them. Only collective bargaining can protect the interests and earnings of workers in firms where executives pocket an outsized share of firms’ incomes.

If collective bargaining became widespread enough throughout the corporate sector, it would likely not only boost workers’ wages but also rein in excess executive pay broadly—actually helping shareholders at the same time. The research on this question is encouraging: Unions and collective bargaining have been found to significantly restrain excess CEO pay (Huang et al. 2017; Nanda, Nishikawa, and Prevost 2025; DiNardo, Hallock, and Pischke 2000). But even if expanded collective bargaining does not tamp down the CEO-to-worker pay ratio beneath our 100:1 threshold, it will at a minimum protect workers’ interests from excess executive power—without foreclosing any policy route shareholders take to restrain excess CEO pay.

Currently, a number of firms with at least some degree of unionization among their workforces have CEO-to-worker pay ratios that exceed 100:1. But this is not an argument against using this threshold as an indicator of dysfunctional corporate governance requiring a strong policy solution. For many of these firms, union coverage among their rank-and-file workforce is far from complete, so this proposal would help fill in glaring gaps. Most of these firms clearly do pay wages that are among the highest in their industries, which signals that the policy of default collective bargaining to protect workers’ interests in the face of dysfunctional corporate governance would largely work. Finally, one key empirical driver of high CEO pay is the simple size of the firm: The CEO-to- worker pay ratio rises steeply with the size of the firm, all else equal.27 The firms with CEO-to-worker pay ratios above 100:1 and some degree of unionization among their workforces are large firms; it would take a very strong effect of collective bargaining on CEO pay specifically to have kept pay ratios at these large firms well below average.

In our world with only 10% union density (and less than this in the private sector), it should hardly be a shock that collective bargaining by itself has not been able to stem what has been an irresistible tide toward high CEO pay. Even large increases in executive pay are not enormous when compared with overall returns to shareholders. Given the strong policy and economic barriers to shareholder activism in restraining CEO pay, it takes truly outrageous pay scales to mobilize enough organized opposition from shareholders. In those firms where it is relatively strong, collective bargaining has protected workers’ interests, and that is the primary goal we are looking to achieve with this policy. The fact that in these firms the CEOs have felt free to shift income to themselves at the expense of other stakeholders (like the shareholders of the firm) is not a sign that default collective bargaining is a bad solution for protecting workers’ interests—it is only a sign that the problem of excess CEO pay needs to be addressed from other angles as well.

Tripling union density is what workers want—and it is entirely achievable

Restoring union density to 1950s levels is not a nostalgic pipe dream. Federal and state policymakers have a clear roadmap for tripling union density. While no single reform discussed in this report is a silver bullet for the problems facing the country, taken together they would address one of the worst: the erosion of workers’ right to bargain for their fair share of the wealth they produce. These reforms are meaningful steps toward tripling union density, and toward an economy where productivity gains are shared broadly, not captured by corporate elites and the superrich.

The evidence in this report shows what is at stake. Tripling union density would deliver nearly $270,000 in extra earnings over the median worker’s career and shift $1.2 trillion annually to working people—reversing a third of the rise in inequality since 1979. Rebuilding union power would narrow racial wage gaps by more than a third, and removing state anti-union policies would help dismantle the legacy of Jim Crow-era campaigns aiming to suppress multiracial organizing. States with higher union density deliver better economic and personal well-being outcomes to their residents, and they have healthier democracies with fewer voting restrictions. At a time when the voting rights of all citizens are under attack, rebuilding union power is a democratic imperative.

Acknowledgments

The authors thank Hilary Wething for her contribution to this report’s section on public education spending. We also thank Matthew Wich, Summer Labor Fellow from the Institute for Social Concerns at the University of Notre Dame, for his support of the project.

Appendix

Methodology

How we measure union density

The Bureau of Labor Statistics provides data on both union membership—workers who are full-fledged union members—and union representation, which includes both union members and workers who are not members but are covered by a collective bargaining agreement. As a result, the share of workers represented by a union is higher than the share of union members. For example, in 2025, 11.2% of workers were represented by a union, but 10% were union members.

In this report, we measure union density as the share of workers who are members of a union. Throughout this report, the terms “union density” and “unionization” refer to those who are members of a union. When measuring union density by state, data are averaged over three years (2023 to 2025) to give a more accurate estimate of state unionization rates and avoid temporary single-year changes due to small sample sizes in some states.

Estimating wage effects of tripling union density

To assess how much higher wages would be after tripling union density, we use state-level annual data on wages and unionization from 1979 to 2019 and compare median wages with unionization rates across states and over time. Specifically, we regress the state median wage on the state union membership rate, controlling for permanent differences in state characteristics, national annual shocks (like recessions), and annual state characteristics like state unemployment rate, minimum wage level, and education, race, age, gender, broad industry, and managerial shares.

Using state-level data from the Current Population Survey Outgoing Rotation Group for 1979–2019, we regress the log median wage on union density with state and year fixed effects and additional covariates: the log minimum wage, unemployment rate, and shares of those with less than a high school degree and just a high school degree, age group shares, and shares of those who are white, Black, Hispanic, married, and shares of those in the public sector, manufacturing, construction, services, and a managerial occupation. Regressions are weighted by state employment levels averaged over 1979–2019. The coefficient on union membership is 0.724 with a standard error, accounting for clustering at the state level, of 0.191.

To calculate effects of increasing union density on racial wage gaps, we use the same wage regression as before, except where the dependent variable is the state annual racial wage gap.

The calculation yielding the $1.2 trillion estimate of the annual increase in workers’ pay from tripling union density is as follows. First, the share of total labor income claimed by the bottom 80% is calculated from data provided by the Congressional Budget Office’s (CBO 2026) Distribution of Household Income data (we focus on the bottom 80% because research has generally identified the positive effect of unions on boosting pay as applying mostly to union and nonunion workers in the bottom 80% of the U.S. workforce). In 1979 this share was 60.5%, but by 2019 it had fallen to 50.1%. We use 50% for our calculations. Then we multiply the 14.5% boost to pay stemming from a tripling of union density by this 50% of labor income to get an estimate (7.25%) of how much total labor compensation would rise in the event of a tripling of union density. In the first quarter of 2026, the National Income and Product Accounts (NIPA) Table 2.1 from the Bureau of Economic Analysis (2026a) reports that total labor compensation was $16.1 trillion. Multiplying this by the 7.25% pay boost from tripling union density yields $1.16 trillion in higher pay for the bottom 80%. We believe this is the appropriate estimate for approximating how much higher wages would be for most workers if union density tripled. While potential caveats merit consideration, none, in our view, hold much force. For example, the regressions used earlier in the paper use median wages, not median total labor compensation, as the dependent variable. However, a long research literature has highlighted that benefit premiums stemming from unionization are likely even larger than wage premiums (Knepper 2020). To the degree that tripling union density works by giving all workers more leverage and bargaining power in labor markets, it seems clear that some of this leverage will be used to obtain broader and more generous coverage of benefits, not just higher wages. Another objection could be that union effects (both direct and indirect effects) are higher in the middle of the wage distribution than in either the lower or upper parts of the distribution, and hence the full 14.5% wage boost would only apply to (say) the second and middle fifths of workers while workers in the lowest and fourth fifth would see smaller wage boosts. However, in work that has estimated both the overall average effect of unionization on nonunion wages and the effect by wage percentile, the overall average effect is roughly one-half the effect at the median, a result that would be consistent with applying the 14.5% premium to half of all labor income, as we do (Fortin, Lemieux, and Lloyd 2021). Other research has similarly found that union effects at the median are very close to overall average effects (Baker et al. 2026).

For our analysis of how wage effects differ across union density levels, we use the same regression of log median wages on union density described above, modified to let the slope differ below and above 15% density using a piecewise-linear specification interacting union density with an indicator for density of at least 15%, while retaining the same controls and state and year fixed effects, weighted by state employment. The difference in slopes is statistically significant at the 1% level.

The scatterplot in Figure B uses 50 state-year bins to show this nonlinearity directly. The binned scatterplot groups the state-year observations into 50 employment-weighted bins by union density and plots each bin’s average real log median wage against its average union density, after partialing out the same controls and state and year fixed effects used in the regression.

State union density groupings

We divided the 50 U.S. states plus the District of Columbia into three equally sized groups based on their current (2023–2025) level of union density.

These groupings are shown in Appendix Table 1. We refer to the 17 states with the highest union density as “high-union-density states” (with 11.9%–25.1% union membership rates); the next 17 states (including D.C.) are “medium-union-density states” (6.8%–11.6%); and the remaining 17 are “low-union-density states” (2.5%–6.1%).

Appendix Table 1

Estimating union density effects of establishing a public-sector duty to bargain and repealing “right-to-work” laws

To estimate the effect of these two policy changes—1) requiring state and local governments to bargain with public-sector workers who choose to unionize, and 2) repealing RTW—we use the CPS Outgoing Rotation Groups for 2023–2025 to compute union density rates in states with stronger union policies (i.e., states with public sector collective bargaining and/or without RTW). We do this separately by industry and major occupation group for the private sector and by level of government (federal, state, and local) for the public sector. We then apply those rates to the workforce of each state that lacks the corresponding policy, holding each state’s industry, occupation, and government-employment mix fixed. Nevada is an exception on the private-sector side: Although Nevada is a RTW state, its private-sector union density is already comparatively high, so we hold its private-sector density at its current level.

Notes

1. “Union membership” refers to workers who are full-fledged union members, while “union representation” includes both union members and workers covered by a collective bargaining agreement but not members. The share of workers represented by a union is thus higher than the share of union members. In 2025, for example, 11.2% were represented by a union and 10% were union members.

2. In 2025, for the fifth consecutive year in a row, approval of unions reached record high levels last seen in the late 1950s, when union membership was roughly triple what it is now.

3. The productivity-pay gap is a measure of how much income is generated in an average hour of work.

4. “Union wage premium” refers to the additional wages paid to union members compared with nonunion workers with similar characteristics.

5. For example, when the top marginal tax rate was 91%, as it was in the 1950s and early 1960s, executives kept far less of each additional dollar of pay than they do under today’s 37% top rate, which reduced the payoff to pursuing ever-larger compensation packages.

6. Relatedly, expanded unemployment benefits during the pandemic boosted workers’ bargaining power and compelled employers to make higher wage offers (Bivens and Banerjee 2023).

7. See appendix for details about our methodology.

8. The real median wage is the wage of the person in the middle of the overall wage distribution, including both union and nonunion workers.

9. The median monthly cost of a mortgage in the U.S. is $1,521 according to Census (2025). Over the course of a year, that is $18,252.

10. See appendix for details about the calculation of this figure.

11. In 2020 and 2021, Congress authorized Economic Impact Payments—better known as the “COVID checks.” Those three rounds of payments added roughly $430 billion to U.S. families’ personal income in those years and were widely viewed as an historically ambitious economic intervention. Data on how much these checks added to personal income can be found at Bureau of Economic Analysis (2026c).

12. Data on federal defense spending in 2025 is from Bureau of Economic Analysis (2026b).

13. Data on the income shares by percentile come from CBO (2026), which contains data through 2022. The calculation of how much the $1.2 trillion gain to wages would reverse the rise in income shares of the top quintile uses data from BEA (2026a).

14. See appendix for details on how we estimated these wage effects.

15. See appendix for details on how we calculated the effects of increasing union density on racial wage gaps.

16. See Appendix Table 1, which shows how we grouped states into low-, medium-, and high-union-density categories.

17. For details on the data and methodology used to produce these estimates, see appendix.

18. For this state-level analysis, we use 2023–2025 averages of union density across states, which is 9.9%, rather than the 10% for 2025 cited earlier in this report.

19. See passage of Protecting America’s Workforce Act and Faster Labor Contracts Act.

20. The last time a major labor law reform was considered was in 2009 with the Employee Free Choice Act.

21. The Protecting the Right to Organize Act has two Republican co-sponsors, the Public Service Freedom to Negotiate Act has six Republican co-sponsors, and the Protecting America’s Workforce Act has nine Republican co-sponsors.

22. One way to determine the exact COLA amount is to use nonseasonally adjusted values of the national Consumer Price Index for All Urban Consumers (CPI-U) published by the Bureau of Labor Statistics. These price index values are easily available, less subject to revision, and widely covered in the media. The simplest COLA adjustment would be to use the annual rate of inflation measured as the percent change between the two most recent annual averages of nonseasonally adjusted CPI-U values. See BLS (2023) for other considerations.

23. CBO estimated CPI in 2026 at 2.8%, 2027 at 2.4%, and then 2.3% after—however those estimates were done before the U.S. war in Iran that increased energy prices, so we estimate COLA at 3% for purposes of the proposal. To calculate full-time, full-year worker wages, we used 2025 average production, nonsupervisory wage rates (EPI 2026f).

24. We specifically would recommend calculating the CEO-to-typical-worker pay ratio using the methodology detailed in EPI (2025). 

25. In order to appear on the ballot, an individual or organization would have to demonstrate a record of collective bargaining over the preceding three years or obtain 100 signatures or 10% of the workforce, whichever is less.

26. The showing of extraordinary circumstances is referenced in the acute care hospital rule: 29 CFR § 103.30 (1989).

27. This correlation between simple firm size and CEO pay is yet another piece of evidence that CEO pay is not about the skills or acumen of individual CEOs, but is driven by noncompetitive market structures and labor markets for executives.

References

AFL-CIO Tech Institute. 2026. “It’s Better in a Union!” (web page). Accessed July 7, 2026.

Ahlquist, John S., Jake Grumbach, and Thomas Kochan. 2024. The Rise of the ‘Union Curious’: Support for Unionization Among America’s Frontline Workers. Economic Policy Institute, July 2024.

American Compass. 2025. PRO-Tip: Only Some Labor Reforms Are Pro-Worker. May 1, 2025. 

Andrias, Kate, and Alexander Hertel-Fernandez. 2026. Democratic Abundance: An Abundance That Works for Workers. Roosevelt Institute, April 2026.

Baker, Michael, Yosh Halberstam, Kory Kroft, Alexandre Mas, Derek Messacar. 2026. The Impact of Unions on the Wage Distribution: Evidence from Higher Education. American Economic Review: Insights 8, no. 2 (June): 196–213.

Biasi, Barbara, and Heather Sarsons. 2022. “Flexible Wages, Bargaining, and the Gender Gap.” Quarterly Journal of Economics 137, no. 1 (February): 215–266. https://doi.org/10.1093/qje/qjab026.

Bivens, Josh. 2026. “Employer Assessment Fees Are Not an Adequate Solution to Low Wages and Large Safety Net Cuts.” Working Economics Blog (Economic Policy Institute), February 27, 2026.

Bivens, Josh, and Asha Banerjee. 2023. Lessons from the Inflation of 2021–202(?). Economic Policy Institute, April 2023.

Bivens, Josh, Celine McNicholas, Kyle K. Moore, and Margaret Poydock. 2023. “Unions Promote Racial Equity” (fact sheet). Economic Policy Institute, July 31, 2023. 

Bondy, Assaf S. 2025. “Workers For Democracy? Trade Unions and the Struggle Against Democratic Backsliding.” Democratization 33, no. 2: 344–368. https://doi.org/10.1080/13510347.2025.2507359. 

Brenan, Megan. 2025. “Labor Union Approval Relatively Steady at 68% in U.S.” Gallup, August 28, 2025. 

Bureau of Economic Analysis. 2026a. National Income and Product Accounts Table 2.1. – Personal Income and Its Disposition. Accessed June 25, 2026.

Bureau of Economic Analysis. 2026b. National Income and Product Accounts Table 1.1.5 – Gross Domestic Product. Accessed June 25, 2026.

Bureau of Economic Analysis. 2026c. COVID-19 and Recovery Archive – Personal Income. Accessed June 25, 2026.

Bureau of Labor Statistics. 2023. How to Use the Consumer Price Index for Escalation. June 8, 2023.

Bureau of Labor Statistics. 2026. Employee Benefits in the United States. Table 2. Medical care benefits: Access, participation, and take-up rates, March 2025. Last modified September 25, 2025. Accessed May 4, 2026. 

Census Bureau. 2025. “Monthly Mortgage Payments for Homeowners Who Moved in 2024 Were $2,225, Above the U.S. Median of $1,521.” September 11, 2025.

Center on Budget and Policy Priorities (CBPP). 2020. The Far-Reaching Benefits of the Affordable Care Act’s Medicaid Expansion (chart book). October 21, 2020. 

Childers, Chandra. 2023. Rooted in Racism and Economic Exploitation. Economic Policy Institute, October 2023.

Clegg, Daniel, Elke Heins, and Philip Rathgeb. 2022. “Unemployment Benefit Governance, Trade Unions And Outsider Protection In Conservative Welfare States.” Transfer: European Review of Labour and Research 28, no. 2: 195–210. 
https://doi.org/10.1177/10242589221094240. 

Congressional Budget Office (CBO). 2026. Distribution of Household Income 2022.

Dean, Adam, Jamie McCallum, and Jake Grumbach. 2023. Labor Unions and the Defense of American Democracy. Economic Policy Institute, December 2023. 

DiNardo, John, Kevin F. Hallock, and Jörn-Steffen Pischke. 2000. “Unions and the Labor Market for Managers.” IZA Discussion Paper, August 2000.

Dixon, Marc. 2007. “Limiting Labor: Business Political Mobilization and Union Setback in the States.” Journal of Policy History 19, no 3: 313–344. https://doi.org/10.1353/jph.2007.0015.

Dresser, Laura, Joel Rogers, and Leslie Vasquez. 2025. The State of Working Wisconsin 2025. High Road Strategy Center, August 2025.

Economic Policy Institute (EPI). 2021. “Unions Help Reduce Disparities and Strengthen Our Democracy” (fact sheet). April 23, 2021.

Economic Policy Institute (EPI). 2025. “CEO Pay Has Skyrocketed Since 1978: Discussion of our Methods” (web page). Last updated September 25, 2025.

Economic Policy Institute (EPI). 2026a. “Average Real Annual Wage (2023$),” State of Working America Data Library. Accessed June 30, 2026.

Economic Policy Institute (EPI). 2026b. “Nominal Wage Tracker” (web page). Accessed June 30, 2026.

Economic Policy Institute (EPI). 2026c. “The Productivity–Pay Gap” (web page). Last updated March 23, 2026.

Economic Policy Institute (EPI). 2026d. “Union Wage Premium – Union Wage Premium, Average (Regression-Based),” State of Working America Data Library. Accessed May 4, 2026.

Economic Policy Institute (EPI). 2026e. “Wage Calculator” (web page). Accessed May 4, 2026.

Economic Policy Institute (EPI). 2026f. “Hourly Earnings by Industry – Production and Nonsupervisory Employees, Average Nominal Wage,” State of Working America Data Library. Accessed June 30, 2026.

Farber, Henry S., Daniel Herbst, Ilyana Kuziemko, and Suresh Naidu. 2021. “Unions and Inequality over the Twentieth Century: New Evidence from Survey Data.” The Quarterly Journal of Economics 136, 3 (August): 1325–1385.

Feigenbaum, James, Alexander Hertel-Fernandez, and Vanessa Williamson. 2018. “From the Bargaining Table to the Ballot Box: Political Effects of Right to Work Laws.” National Bureau of Economic Research Working Paper no. 24259, January 2018.
https://doi.org/10.3386/w24259.

Fortin, Nicole, Thomas Lemieux, and Neil Lloyd. 2021. Labor Market Institutions and the Distribution of Wages: The Role of Spillover Effects. Journal of Labor Economics 39, no. S2.

Fortin, Nicole, Thomas Lemieux, and Neil Lloyd. 2022. “Right-to-Work Laws, Unionization, and Wage Setting.” National Bureau of Economic Research Working Paper no. 30098, June 2022. https://doi.org/10.3386/w30098.

Freeman, Richard B., and Casey Ichniowski. 1988. When Public Sector Workers Unionize. Chicago, Il.: Univ. of Chicago Press.

García, Emma, and Eunice Han. 2021. The Impact of Changes in Public-Sector Bargaining Laws on Districts’ Spending on Teacher Compensation. Economic Policy Institute, April 2021.

Glass, Aurelia. 2025. “Everybody Likes Unions.” Center for American Progress, November 4, 2025. 

Hertel-Fernandez, Alexander, and Alix Gould-Werth. 2020. Labor Organizations and Unemployment Insurance: A Virtuous Circle Supporting U.S. Workers’ Voices and Reducing Disparities in Benefits. Washington Center for Equitable Growth, October 2020.

Huang, Qianqian, Feng Jiang, Erik Lie, and Tingting Que. 2017. “The Effect of Labor Unions on CEO Compensation.” Journal of Financial and Quantitative Analysis 52, no. 2: 553–582. https://doi.org/10.1017/S0022109017000072.

Jackson, C. Kirabo, Rucker C. Johnson, and Claudia Persico. 2016. “The Effects of School Spending on Educational and Economic Outcomes: Evidence from School Finance Reforms.” The Quarterly Journal of Economics 131, no. 1: 157–218. https://doi.org/10.1093/qje/qjv036. 

Jackson, C. Kirabo, and Claire L. Mackevicius. 2023. “What Impacts Can We Expect from School Spending Policy? Evidence from Evaluations in the United States.” American Economic Journal: Applied Economics 16, no. 1: 412–446. https://doi.org/10.1257/app.20220279.

Johnson, Theodore R., and Max Feldman. 2020. The New Voter Suppression. Brennan Center for Justice, January 2020. 

Kaufman, Dan. 2018. “A Labor Day Reflection on Unions, Race, and Division.” The New Yorker, September 3, 2018.

KFF. 2026. Health Insurance Coverage of Population Ages 0-64: 2024. Accessed May 4, 2026. 

Knepper, Matthew. 2020. “From the Fringe to the Fore: Labor Unions and Employee Compensation.” The Review of Economics and Statistics 102, no. 1: 98–112.

Ma, Jennifer, Matea Pender, and Xiaowen Hu. 2025. Trends in College Pricing and Student Aid 2025, College Board.

McNicholas, Celine, Margaret Poydock, and Lynn Rhinehart. 2021. “Why Workers Need the Protecting the Right to Organize Act” (fact sheet). Economic Policy Institute, February 9, 2021.

McNicholas, Celine, Margaret Poydock, and Heidi Shierholz. 2026. Workers’ Resolve Drives Increase in Unionization in 2025. Economic Policy Institute, February 2026.

McNicholas, Celine, Margaret Poydock, Heidi Shierholz, and Hilary Wething. 2025. Unions Aren’t Just Good for Workers—They Also Benefit Communities and Democracy. Economic Policy Institute, August 2025.

Nack, David, Michael Childers, Alexia Kulwiec, and Armando Ibarra. 2019. “The Recent Evolution of Wisconsin Public Worker Unionism Since Act 10.” Labor Studies Journal 45, no. 2: 147–165. https://doi.org/10.1177/0160449X19860585.

Nanda, Vikram, Takeshi Nishikawa, Andrew Prevost. 2025. “The Impact of Unions on Compensation Consultants and CEO Pay.” Financial Management 54, 89–122. 
https://doi.org/10.1111/fima.12472.

Pierce, Michael. 2017. The Origins of Right-to-Work: Vance Muse, Anti-Semitism, and the Maintenance of Jim Crow Labor Relations. The Labor and Working-Class History Association, January 12, 2017.

Rosenfeld, Jake, Patrick Denice, and Jennifer Laird. 2016. Union Decline Lowers Wages of Nonunion Workers. Economic Policy Institute, August 2016. 

Securities and Exchange Commission. 2007. “Executive Compensation” (web page). Last modified January 4, 2007.

Sherer, Jennifer. 2026. Holding the Line: Rights to Unionize and Collectively Bargain. Economic Policy Institute, February 2026.

Sherer, Jennifer, and Elise Gould. 2024. “Data Show Anti-Union ‘Right-to-Work’ Laws Damage State Economies.” Working Economics Blog (Economic Policy Institute), February 13, 2024.

Sherer, Jennifer, and Monique Morrissey. 2026. Stronger Collective Bargaining Laws Will Benefit All Virginians. Economic Policy Institute, January 2026.

Shierholz, Heidi. 2024. “Middle-Out Economics Is Good for Workers, Their Families, and the Broader Economy.” Working Economics Blog (Economic Policy Institute), March 19, 2024.

Sojourner, Aaron, and Adam Reich. 2025. “Americans Favor Labor Unions Over Big Business Now More Than Ever.” Working Economics Blog (Economic Policy Institute), May 20, 2025. 

Stelzner, Mark, Eric Hoyt, and Toushita Ramchurn. 2019. Structured Conflict: Changes in Federal and State Labor Laws and Strike Activity, 1950 to 2017. Political Economy Research Institute (PERI), University of Massachusetts Amherst, May 2019.

United States Department of Agriculture (USDA). 2017. “2015 Expenditures on Children By Families” (web page). Accessed June 30, 2026.

Van Green, Ted. 2025. “Majorities of Adults See Decline of Union Membership as Bad for the U.S. and Working People.” Pew Research Center, August 27, 2025. 

Zhang, Elizabeth. 2026. Nearly 3 Million Uninsured Adults Would Gain a Path to Medicaid Coverage if Their States Adopted ACA Medicaid Expansion. Center on Budget and Policy Priorities, March 19, 2026.

Defenders of the OBBBA cuts claim that, because a large share of these cuts result from introducing new bureaucracy around work-reporting requirements for “able-bodied adults without dependents” (ABAWDs), they will not fall on more vulnerable populations like children or retirees. This is not true. The OBBBA has many near-direct cuts to vulnerable populations’ participation in these programs. More importantly, these cuts will have large spillover effects through families and communities that will harm vulnerable populations—including children.

This post highlights how important Medicaid and SNAP spending is to children, with a particular focus on how these programs support public education. It then outlines some ways that the legislated cuts in the OBBBA will damage this support, either directly or through clear spillover effects through damage to local economies.

The importance of Medicaid and SNAP to children and schooling

Around 40% of school-age children are covered by Medicaid for health insurance, and nearly 30% of school-age children live in families that receive SNAP benefits. Medicaid funding is key for public education. It provides an estimated $7.5 billion annually to public schools to fund essential health and learning and development services. As a share of total K–12 student revenue, Medicaid provides up to 2% of school funding to some states and up to $325 per pupil in revenue in some states.

Medicaid is particularly crucial for special education. Public schools are required by law to provide appropriate care and services for individualized education plans, and Medicaid is a big source of funding for these plans. In a recent survey of school districts, 86% reported they use Medicaid funds for school health staff salaries, and 59% use the funds for contracted services for mental and behavioral health. When asked what they would have to cut if they lost Medicaid funding, 8 in 10 school district staff predicted reductions in school health personnel and services.

How Medicaid and SNAP cuts will still harm children through direct fiscal pressures

While the OBBBA cuts to Medicaid and SNAP do not directly harm children through cuts in children’s health insurance or their food stamp eligibility, changes to eligibility criteria and cost pressures on state budgets will make it onerous to maintain current spending levels for one of the main public institutions that support children: public schools. For example, if states want to maintain coverage for any ABAWDs who lose coverage because of the work requirements stipulated in the OBBBA, states will have to come up with funding for it themselves.  Additionally, the SNAP legislation requires that states now pay for 25% of their administrative costs that previously were fully funded from federal sources. With less federal money, states will either have to cut spending elsewhere in their budgets or (preferably, but politically difficult) raise revenue. None of these are easy political choices, and they all have some spillover effects. As one of the largest budget items in states, K–12 education funding will likely be squeezed in any state looking to maintain any of the services cut in the OBBBA without raising revenue.

How Medicaid and SNAP cuts will harm children through damage to local economies

Beyond these direct budgetary impacts, the OBBBA Medicaid and SNAP cuts could also reduce purchasing power in local economies and put upward pressure on unemployment rates, which could reduce economic security and opportunity for children in those areas.  Adults who receive Medicaid and SNAP don’t have to use their household budget to pay for health care or (some) groceries. Medicaid and SNAP allow them to maintain access to these necessities while also purchasing other goods in their local economy, such as electricity, child care, or rent. When these programs are cut, people face higher health and food costs, which can force them to cut back on purchases and reduce demand throughout the local economy.  

Moreover, areas particularly dependent on Medicaid and SNAP funding are least likely to be able to weather a reduction in aggregate demand for goods and services. Figure A shows that there is a clear positive relationship between a county’s unemployment rate and the income from Medicaid and SNAP. In other words, counties with high unemployment levels are also counties where a large share of their total income comes from Medicaid and SNAP funding.  

To be fair, at the national level, it is possible that the reduction in spending caused by Medicaid and SNAP cuts will be neutralized by policies that raise incomes and demand elsewhere (say, by the tax cut provisions of the OBBBA or a Federal Reserve interest rate cut). But areas that are already facing high unemployment are missing out on positive trends in the U.S. economy. For these counties, the demand shock caused by Medicaid and SNAP cuts could be large, leading to job loss in these counties.

Damage to local economies hurts students and workers

Job loss from the spillover effects of the OBBBA cuts to local economies will create stress and financial insecurity for families, a catastrophe that research has shown can harm student achievement.

Parental job loss plays a key role in reducing parent and child well-being  and has been shown to  reduce a child’s likelihood of grade completion. Moreover, parental job loss is associated with a lower likelihood of obtaining postsecondary education, particularly in Black families. Causal evidence on community level job losses found a reduction in test scores for students, following the closing of a business. The effects were stronger for older children and for children in lower-income households. Even if the Medicaid cuts mostly target adults (ABAWDS), the reduction in local aggregate demand stemming from these funding cuts will hurt all families in a given region, and particularly for children, stand to harm their educational achievement. 

Conclusion

The cuts to Medicaid and SNAP mandated by the OBBBA will harm children. Medicaid and SNAP currently fund key health and food services for public education, and while these functions aren’t directly threatened under the OBBBA, the cost pressures states will face from federal funding cuts will threaten the ability of these programs to serve students. Moreover, there are clear spillover effects through damage to local economies. By reducing the amount of available income in these areas, Medicaid and SNAP cuts run the risk of imposing a sharp anti-stimulus to local economies. This effect could be large in demand-constrained counties (studies on the multiplier effect of Medicaid consistently show it has large effects on spending).   The claim that OBBBA won’t harm children is patently wrong, and the harms from it will plague children for a long time to come.

In the last two years, Republicans in Tennessee have attempted to push legislation that would violate Plyler to set the stage to challenge the court decision (although neither proposal passed). Last year, the state proposed charging undocumented students tuition for public schools, and this year, Tennessee attempted to pass legislation to track the immigration status of all public school students.

The 1982 ruling of Plyler v. Doe is notable because it stated that the harm of not educating undocumented children would be worse for society than providing a basic education to all children in the U.S. The ruling recognized the huge positive spillovers public education has on the U.S. labor market, public health, and civil society and that leaving immigrant children out of public education would create an “underclass” in U.S. society.

Moreover, if the move to deny public education to children in the U.S. is successful, particularly in pockets of the country where immigrant children are a substantial share of the student population, it will lead to an extraordinarily high cost for the students who remain in public school.

Some school costs are hard to adjust, regardless of the number of students

Across the country, an estimated 17% of school-aged children live with at least one noncitizen adult, according to the Kaiser Family Foundation. If these students were to leave suddenly, schools would be left to educate a fewer number of kids without any time to adjust their fixed costs. For example, when Alabama passed an immigration data collection law, more than 13% of Hispanic schoolchildren withdrew from classes.

At first, it would seem that reducing enrollment would reduce both total revenue and the number of students needing educational services proportionately, which should leave the schools’ ability to provide education unaffected. But schools can’t adjust every educational cost quickly: School bus routes still need to circulate to all stops, even if there is one fewer child in need of transit; buildings need to be heated and cooled, even if classroom size goes down; and guidance counselors and support staff are still required to support the remaining students.

School districts will pay more per pupil if student enrollment declines because of immigration enforcement

Since these fixed costs can’t be adjusted in the short run, when total revenue declines due to families’ fears of deportation, districts are stuck paying more per pupil on costs they can’t adjust. Effectively, native-born students who remain in public schools receive no additional benefit when immigrant students are denied services and rights. Districts instead will be paying more on costs that can’t be adjusted and getting less on the costs that can be adjusted for fewer students.

Calculating the cost under two different scenarios

We call all the costs of downward adjustment that occur when enrollment is reduced the fiscal externality. This means the per-pupil funds each district would require to maintain the same level of spending for remaining public school students due to a rapid decline in enrollment. This cost is entirely borne by state and local education budgets and leaves districts unable to deliver the same level of instruction to the remaining public-school pupils.

For districts with a large share of school-aged children in immigrant families, the costs of losing these students could be substantial. Table 1 shows the top-25 school districts, based on the number of K–12 students that are in immigrant families and the corresponding fiscal externality for two scenarios. In the first scenario, half of these students stop attending public school (referred to as the lower bound in the table). In the second scenario, all of these students stop attending public school (referred to as the upper bound in the table).

<

The implications of a reversal of Plyler v. Doe or any type of policy restricting immigrant children’s access to public education would be extreme for these districts. In Houston, Texas, where 62% of students in the school district are Hispanic, we estimate that nearly 63,000 students may live in a household with immigrants and as such, might be vulnerable to dropping out of school due to anti-immigration efforts. The lower bound shows the costs if half of these students stopped showing up. Houston School District would have to reduce services by $1,654 for each remaining (and disproportionately native-born) public school student. This decline translates to a total fiscal externality of $268 million a year, or 11% of the total budget for the school district.

The extraordinarily high cost to native-born students of losing students in immigrant families due to anti-immigrant policies highlights the hypocrisy in the anti-immigrant movement. If all students in immigrant families stopped attending public school tomorrow, not only would those students suffer from the lack of public education, but the quality of public education would be much worse for the remaining students in those same schools. In short, no one wins. When the costs are tallied up, it’s clear that these policies are not about improving education quality for native-born students, but instead are malicious attacks on the institution of public education in the U.S.

Key takeaways:

• The vast majority (85%) of young college graduates work in occupations that have seen strong employment growth in recent years.

• Young college graduates, like college graduates in general, are more likely to work in AI-exposed occupations than the overall workforce—and considerably more likely than young noncollege workers.
• But both young college graduates and young noncollege workers have experienced rising unemployment over the last three years, suggesting AI is not likely to be driving labor market weakness.

In the first blog post of our Class of 2026 series, we showed that the strong labor market for young college graduates of the early 2020s had begun softening in recent years. A growing share of young college graduates are seeking employment, but because their employment rates have not kept up with this job search, their unemployment rate has risen faster than the overall rate. The second blog post in the series discussed the industries where young college graduates worked. We found that recent graduates work in growing industries, but are forced to enter a weakened labor market with less job turnover, deteriorating their ability to break in. Young college graduates work in the tech sector at a similar rate to college graduates, and there is no clear evidence that tech sector employment is significantly decreased despite warnings about the advancement of AI.

In this blog post, we delve deeper into the occupations where young college graduates are likely to work.1 We examine whether it has been relatively more difficult to secure employment in these fields as the labor market has weakened. We also scour the data for signs that exposure to AI-related occupations may disproportionately affect the prospects for young college graduates as they enter the labor market.

Most young college graduates work in occupations with strong growth

Over 60% of young college graduates work in professional and related occupations or management, business, and financial occupations. Figure A displays the share of employment in each occupation or occupation grouping for young college graduates ages 22 to 27, all college graduates, and young workers without a four-year college degree. Occupations in the figure appear in order of the share of young college graduates employed in each, from largest to smallest. Over half (62.8%) of young college graduates work in professional, management, business, and financial occupations. Workers of any age with a college degree are slightly more likely to work in those two occupations (64.5%), though more likely in management occupations than professional occupations. On the other hand, nearly half (48.3%) of young noncollege workers are in service occupations or farming, construction, installation, and production occupations.

Figure A

Figure B shows the change in employment in each occupation between 2019 and 2026 and between 2023 and 2026, arranged in the same order as Figure A for comparison. Since 2019, management, business, and financial occupations and transportation and material moving occupations experienced the most growth, followed by professional and related occupations.

The top four occupations for job growth since 2023 account for 85% of young college graduate employment. The occupations with employment losses over the last three years were more likely to employ young noncollege workers than college graduates. It doesn’t appear that the occupations where young college graduates tend to work have been hit particularly hard in the last couple of years.

Figure B

While there has been job growth among occupations that tend to be filled by young college graduates, some worry about an increase in labor market underutilization, i.e., when workers with a college degree wind up working in jobs that typically don’t require one. Using O*NET data2, the New York Federal Reserve tracks this type of labor market underutilization. While the share of recent college graduates working at a job that doesn’t require a college degree has ticked up slightly over the last three years, it remains lower than it was for workers who graduated in the aftermath of the Great Recession. Even as late as 2017, young college graduates were working at these noncollege jobs at higher rates than they are today.

While college-educated workers are in more AI-exposed occupations, this does not appear to be driving labor market weakness

Much has been written in the last few years about AI exposure and its impact on the labor market. Using data from ADP, a large payroll processing company, Brynjolfsson, Chandar, and Chen find that entry-level workers in AI-exposed occupations—particularly AI uses that automate, not augment their work—have experienced an employment decline larger than that of older workers in the same occupations and all workers in less exposed occupations, explaining some of their stagnant overall employment growth. Atkinson and Yamco also find that declines in AI-exposed occupations are tied to lack of hiring rather than layoffs, hitting harder for young people attempting to enter the labor market. The second blog post in our series noted an across-the-board slowdown in hiring—which hurts the job prospects of all young workers, not only those in the industries most affected by AI.

On the other hand, researchers at the Yale Budget Lab argue that there has only been a slight increase in the shift in the occupation mix of employment, which would be evidence of AI automating jobs. They find that high AI-exposed occupations—determined by the top quintile of AI exposure—have yet to show declining employment, so no “dissimilarity” between young and older college graduates in terms of occupation mix has materialized. Raderman also finds that there isn’t strong evidence that AI is responsible for weaker labor market outcomes for recent college graduates, using evidence from Tillerman on college majors paired with change in unemployment.

Given the variation in assessments, we wanted to take a look at the data ourselves. Gimbel, Kendall, and Kulsakdinun have done an admirable job of summarizing the literature that attempts to classify AI exposure and propose a weighted aggregate measure of AI exposure.3 We employ this measure to investigate whether young college graduates may be more likely to be at risk in AI-exposed occupations than other workers.

In Figure C, we display the AI exposure of occupations weighted by the share of the entire workforce in each occupation. Moving from the left to the right on the figure increases AI intensity. For instance, professional and office & administrative support occupations are more AI exposed (to the right), while production, transportation, and service occupations are less AI exposed (to the left). Overall, the mean AI exposure score is 0.23.4

Figure C

In Figure D, we show the distribution of select demographic groups by occupation and AI exposure. As with earlier analysis, we compare young college graduates with all college graduates and young noncollege workers, in separate panels in the figure.

According to the aggregate measure, college graduates do have higher AI exposure in the labor market than the overall workforce. It is clear there is more mass in the direction of higher exposure (to the right) and their mean exposure is 1.07, higher than that of workers writ large. But the AI exposure of young college graduates isn’t any higher than that of college graduates in general. Mean AI exposure among young college graduates is 1.00.

Figure D

What is striking is that the AI exposure among young college graduates (1.00) is considerably higher than that of young noncollege workers (-0.61). If AI was driving labor market outcomes, we’d expect young college graduates to fare worse in today’s economy, e.g., see larger declines in employment or faster increases in unemployment. But, when we compare unemployment rates as we did in the first blog post of this series, both groups experienced similar increases in unemployment over the last two to three years. Trends in employment rates were also consistent across these groups.

Since the weakening labor market is hitting both young college and noncollege workers alike, it’s hard to argue that AI is uniquely causing job losses for new labor market entrants graduating from college now or in recent years. These findings are consistent with the literature, as there is currently no consensus about the effects of working in AI-exposed occupations on employment thus far.

1. Throughout this brief, we define young college graduates as people between the ages of 22 and 27 with only a four-year college degree. Unlike similar analyses of young workers, we do not exclude young college graduates who are currently enrolled in school, but the results here are robust either way. Unless otherwise noted, data for 2026 represent a 12-month average from April 2025 through March 2026 for the most up to date and reliable estimates, which removes seasonality and increases sample sizes.

2. O*NET or the Occupational Information Network provides the largest up-to-date database of information about workers sorted into detailed occupations. Information provided is about skills, abilities, education, training, and more.

3. We use an updated summary AI exposure PCA score (principal component analysis weighted standardized z-score) provided by the authors, May 13, 2026.

4. The PCA score scale is centered at 0, the unweighted mean across occupations.

Key takeaways

• The depressed overall hires rate is a key driver of new labor market weakness for young college graduates, as it makes it harder for them to break into the labor market. This is true across industries, not just in those that disproportionately employ young college graduates—suggesting the culprit is not a structural change in the economy like AI but a labor market in which employers are hiring less and workers are holding onto the jobs they have.
• The information sector—posited to be more AI-exposed—has experienced recent job losses but employs only 2.3% of young college graduates.
• High-tech industries, which employ about 1 in 10 college workers, expanded at a historically rapid pace in the early 2020s but have shown signs of softening over the last three years.

The early 2020s labor market for young college graduates was strong. But, as we showed in this series’ first blog post, the Class of 2026 is graduating college into a labor market that has notably weakened in the past two years. A growing share of young college graduates are looking for jobs, but their employment rates have not kept pace—meaning unemployment is rising faster for young graduates than for the overall workforce. While their outcomes remain better than those of their noncollege counterparts, the uptick in unemployment has been a rising concern.

In this blog post, we delve deeper into the industries where young college graduates are likely to work,1 examining whether it has been relatively more difficult to secure employment in these fields as the labor market has weakened. Our analysis first examines employment changes, then turns to labor market flows, including hires and separations rates. We also scour the data for signs of contraction in the tech sector that may disproportionately affect the prospects for young college graduates as they enter the labor market.

In the third blog post in the series, we will examine the occupations where young college graduates work with particular attention to occupations that may have grown or shrunk, as well as to those most exposed to AI.

Young college graduates work in industries with strong growth in this business cycle

Over half of young college graduates work in private education and health services, professional and business services, or public-sector jobs. Figure A displays the share of employment in each industrial sector or sector grouping for young college graduates ages 22 to 27, all college graduates, and young workers without a four-year college degree. Industries in the figure appear in order of the share of young college graduates they employ, from largest to smallest.

The types of jobs where young college graduates work look similar to those of college graduates generally. Young workers without a college degree (i.e., noncollege) are far more likely to work in trade, transportation, and utilities; mining, construction, and manufacturing; or leisure and hospitality. All groups of workers are least likely to work in the information sector, closely watched for signs of AI-induced displacement.

Figure A

Figure B shows the change in employment in each sector between 2019 and 2026 and between 2023 and 2026, arranged in the same order as Figure A for comparison. The two fastest growing sectors since the last business cycle peak occurred in the two largest sectors for young college graduates: private education and health services and professional and business services.

Since the rollout of ChatGPT, many have looked at industries and occupations likely to be exposed to AI to see whether this has led to weaker job growth. Among the most closely watched of these industries is the information sector, which has seen an 8.5% employment decline since 2023.2 While these losses are notable—and especially relevant to understanding AI’s fingerprints on the labor market for young college-educated workers—it cannot be overemphasized just how small this sector is in the overall economy. Less than 2% of overall employment is in the information sector, including only 2.3% of young college graduates. Further, the sector saw a rapid employment expansion between 2019 and 2023—the employment loss between 2019 and today is just 2.0%.

Figure B

High-tech sectors have relatively more college graduates but haven’t experienced large AI-induced employment losses

In recent years, the Census Bureau has created an experimental data series on high-tech industries to better understand business dynamics. These include both manufacturing and service sector components of high tech.3 We translate their NAICS classification for high-tech industries into Census Industry Classifications used by the Current Population Survey (CPS) to determine the likelihood of young college graduates working in these sectors.

In the 2026 economy, about 5.6% of workers were employed in what the Census considers high-tech industries. Just about 1 in 10 (9.9%) of the college-educated workforce works in the tech industry. Young college graduates are similarly represented: 10.3% work in tech.

Overall, the Current Establishment Survey tells us that tech industry employment tracked changes in overall private employment in the prior business cycle (between 2007 and 2019) but expanded sharply in the early 2020s and has softened a bit in the last three years. Since 2023, the tech sector has fallen by 0.7%. While overall employment using CPS does show modest growth, neither shows large swings that suggest a large impact for young college graduates.4

Weak hires may be the biggest culprit to labor market weakness for young college graduates

The Job Openings and Labor Turnover Survey (JOLTS) can shed light on the question of whether entry-level workers—with or without college degrees—are facing a harder labor market to break into. While JOLTS doesn’t include demographic characteristics, it presents jobs openings as well as rates of hiring, layoffs, quits, and other separations. Today’s economy has substantially less churn than during the recovery from the pandemic, when millions of workers reentered the labor market after mass layoffs—many quit soon after as they searched for, and generally found, better opportunities.

Figure C shows the hires and separation rates. The lighter colors represent the monthly seasonally adjusted data for each series while the darker colors represent a 12-month moving average that provides a better overall picture of recent trends, smoothing out some data volatility. Over the last five years, the hires rate has steadily fallen and now sits at levels last seen in 2013 and 2014, when the labor market was still struggling to recover from the Great Recession.

The total separations rate includes quits, layoffs and discharges, and other separations.5 As with the hires rate, the separations rate has been declining over the last few years and now sits about where it was in 2014. Much of this is due to reductions in quits. Quits are higher when workers feel confident that they will find better job opportunities. Right now, workers are sitting tight, more so than any point in the past 10 years. Taken together, there is simply less churn in the labor market. But reduced churn is not inherently bad. If the frantic labor market of the early 2020s led to many workers and employers finding satisfactory matches, it could make sense that the following years would see less churn than normal. But for young workers looking to enter the job market, a reduction in hiring can make it harder to find a job.

Figure C

Table 1 breaks down the change in the hires and separations rate over the last three years, again using 12-month moving averages to smooth some volatility in the data. The industries are listed in order of the share of young college graduates they employ, similarly to Figure A. Overall, the hires rate fell 0.8 percentage points and the separations rate fell 0.6 percentage points between 2023 and 2026. The industries where young college graduates are more likely to work saw smaller reductions in both hires and quits than the overall. Industries where young workers without a college degree are more often found—over a quarter are in trade, transportation, and utilities—saw greater losses. Finally, leisure and hospitality, where young noncollege are more than twice as likely to work as young college graduates, saw the largest declines in hiring.

It does not appear that the industries where young college graduates tend to work are experiencing more weakness than other industries. Job gains are just as strong, if not stronger, and hiring hasn’t fallen as far in other industries. In short, there does not seem to be any profound structural change in the economy affecting the industry composition of employment—AI or anything else—that would easily explain the softening of the labor market for young college graduates in recent years. What it does appear to be is a harder labor market for young workers to break into when employers are less likely to hire and workers are more likely to sit tight in the job they have.

1. Throughout this blog post, we define young college graduates as people between the ages of 22 and 27 with only a four-year college degree. Unlike similar analyses of young workers, we do not exclude young college graduates that are currently enrolled in school, but the results here are robust either way. Unless otherwise noted, data for 2026 represent a 12-month average from April 2025 through March 2026 for the most up to date and reliable estimates, which removes seasonality and increases sample sizes.

2. ChatGPT was first introduced in November 2022 but took several months for more widespread usage.

3. High tech industries: Computer and Peripheral Equipment Manufacturing, Communications Equipment Manufacturing, Semiconductor and Other Electronic Component Manufacturing, Navigational, Measuring, Electromedical, and Control Instruments Manufacturing, Aerospace Product and Parts Manufacturing, Software Publishers, Data Processing, Hosting, and Related Services, Other Information Services, Architectural, Engineering, and Related Services, Computer Systems Design and Related Services, and Scientific Research and Development Services.

4.  It’s not unusual for the CPS and CES to display small differences in employment levels or trends considering nontrivial differences in their methodologies. 

5. Other separations include separations due to retirement, death, disability, and transfers to other locations of the same firm.

But diverting public funds away from public K–12 schools and toward private schools does not guarantee educational opportunities will be expanded for all students—and this is especially true in rural communities. Most obviously, because students in rural communities often don’t have a private school option and therefore cannot use the vouchers, state voucher programs—which are financed by all the taxpayers in a state—amount to an education subsidy for wealthy urban families at the expense of strong public schools. Moreover, for rural areas that can support multiple school systems, voucher programs introduce a potentially large cost for the students that remain in public schools, as any sharp drop in public school enrollment will raise the fixed cost per pupil of running schools. For example, school facilities and staff that are efficient for 1,000 students in a school may no longer be efficient if enrollment were to drop to 800 or 900.

Voucher programs work like this: Parents who wish to send their kid to private school can receive public funding to cover part of the tuition or education-related expenses, rather than paying out of pocket. In states with vouchers programs, this added cost to government of paying for private educational expenses makes a big dent in state budgets—see examples here, here, and here. These programs also often entail fraud and abuse of funds and strip away funding for public schools. As a share of K–12 budgets, voucher spending accounted for as much as 26% in 2025, squeezing public schools of sorely needed funds. Moreover, recent reports have documented accounts of voucher funding getting used for high-end concert tickets and rideshare apps like Uber and Lyft. For wealthy parents in urban districts who were already planning to send their kids to private school, these slippery regulations and extra funding for education expenses are a feature, not a bug, of voucher programs. Vouchers are disproportionately taken up by students already attending private school, compared with those who consider a private school option when voucher laws get passed in their state.

For students in rural areas with no private school option, voucher programs simply mean there is less to spend on public schools, which leads to teacher shortages, fewer educational opportunities, and worse building maintenance. In rural communities with homeschooling or private school options, voucher programs impose an added cost to public education when students transition from public to private school.

We call this cost the fiscal externality of voucher programs, and it is borne by school districts, students, and their families when voucher-driven declines in student enrollment intersect with the fixed nature of many school costs. In rural districts, many key education costs—such as interest on bonds issued in the past, heating, electricity for school buildings, bus drivers, and even some staff—cannot easily adjust to student enrollment declines.

While public schools’ fixed costs do not decline when they lose students to voucher programs, their revenue does. Thus, when students in rural areas take up vouchers to leave public school for private school or homeschool, public schools have less revenue to cover the same level of fixed costs. The costs that can be adjusted—such as supplies or certain personnel—will get forced down due to shrinking school budgets. These variable costs are crucial for effectively educating children, meaning students who remain in public schools will pay the price of voucher program takeup.

This fiscal externality therefore leaves districts unable to deliver the same level of instruction to the remaining public school pupils. When students leave public schools in rural areas with voucher programs, there are fewer resources available on a daily basis to educate kids—fewer teachers and other staff members and fewer curriculum and education supplies. Education quality suffers.

How large is the fiscal externality that voucher programs impose on public schools in rural districts? Take the McComb Local School District in Ohio, which had 627 students in 2022 and is classified as a rural district according to the U.S. Census. Using EPI’s Fiscal Externality Calculator, we estimate that a 5% decline in enrollment would lead to an increased cost of $520 per pupil for the remaining students in the district, or a total of $309,530.

The key assumption is that there is some fraction of schools’ costs that is fixed and can’t be adjusted in the near term when enrollment falls. We assume that instruction and services costs (the cost of teachers and services like transportation, counseling, nurses, and school administrators) can only partially adjust to changes in enrollment. Specifically, we assume that when enrollment declines, instruction costs are only able to adjust by 50% of the enrollment decline, and service costs are only able to adjust by 20%. We assume that capital and building and maintenance costs can’t be adjusted at all. (Users can set their own adjustment rates for their school districts using the fiscal externality calculator here. The method behind this calculation is detailed in our report.)

Under these assumptions, aggregating all the rural Ohio districts using the rural categorization of school districts from the National Center for Education Statistics, a voucher-driven 5% enrollment decline would impose a fiscal externality of just over $206 million on Ohio public schools.

Rural districts have the most to lose when states enact voucher programs. For rural communities, vouchers are not a cost-free policy that simply expands education options for children—they are a subsidy for wealthy urban and suburban families at the expense of strong public schools. Voucher programs also introduce a large potential cost for the students that remain in rural public schools. The public spending declines associated with the introduction of vouchers will reliably cause significantly worse educational outcomes at a time when states should be spending more—not less—on public schools. States that promote voucher programs at the expense of funding for strong public education are signaling that rural students are not a priority. 

Key takeaways

• Proposed state legislation to extend full, equal collective bargaining rights to all state and local government workers can help Virginia:
  • reduce the state’s large public-sector pay gap
  • improve public services
  • reduce staff vacancies and turnover
  • decrease racial and gender wage disparities
• Strong collective bargaining rights and increased unionization rates are highly correlated with numerous, widely shared benefits including higher wages, more equitable state economies, and healthier democracies.
• Virginia currently has one of the largest public-sector pay gaps in the nation. State and local government employees in Virginia earn, on average, 26.7% less than private-sector peers with similar education and experience.
• The public-sector pay gap varies across states and is largest in states like Virginia where most public employees lack collective bargaining rights. In Virginia, collective bargaining is currently banned for state employees and only recently became permitted for some local government employees.

In 2026, Virginia lawmakers are poised to consider transformative legislation that would extend full collective bargaining rights to public employees at all levels of state and local government. The benefits of comprehensive collective bargaining rights would extend far beyond affected public employees who would enjoy better working conditions. Stronger state collective bargaining laws can help improve the quality of public services and economic outcomes for all Virginians.

Data show that strong collective bargaining laws help states address persistent public-sector pay gaps, reduce staff vacancies and turnover, and lead to higher unionization rates (Morrissey and Sherer 2024). Increased unionization rates are highly correlated with numerous, widely shared benefits including more equitable state economies and healthier democracies (McNicholas et al. 2025).

For Virginia, expanding public-sector bargaining is a critical next step in building an economy that works for all. The expansion will reverse a long history of anti-worker state policies that have suppressed wages and limited workers’ power in the labor market by blocking pathways to unionization. Such policies have resulted in greater income inequality and persistent racial and gender wage disparities (Bivens and Shierholz 2018; Mishel and Bivens 2021). Strong, comprehensive state legislation covering public employees’ labor rights is also especially important at a moment when the federal government has been attacking the jobs, working conditions, and union contracts of over 235,000 federal civil servants residing in Virginia, and threatening long-standing federal protections of all workers’ rights (EPI 2025b; Oakford and Poydock 2025). 

This report examines how public-sector workers with limited or no collective bargaining rights fare compared with public-sector workers with well-established collective bargaining rights. To do so, we estimate pay differences between state and local government workers and private-sector workers with similar education and experience. In this analysis, Virginia is among a minority of states in which state employees have no bargaining rights and only some local government employees have limited bargaining rights. By contrast, 27 states have well-established collective bargaining rights for state and local government workers (Table 1).

The right to bargain collectively over pay is associated with higher unionization rates (union membership as a share of the workforce) in a given state. This report shows that collective bargaining rights and union strength help state and local government workers narrow the pay gap with private-sector workers. We show that the pay gap for public employees is significantly larger when these workers have weak or no bargaining rights, like public employees in Virginia, which has one of the largest public-sector pay gaps in the nation (-26.7%).

Virginia public employees lack collective bargaining rights

Proposed legislation in Virginia would, for the first time in the state’s history, guarantee that all state and local government employees enjoy labor rights similar to those of their private-sector peers (whose rights to collectively bargain are covered under the federal National Labor Relations Act) (HB 1263 and SB 378). The Virginia Assembly passed similar legislation (HB 2764 and SB 917) in 2025, only to have it vetoed by then-Governor Glenn Youngkin (McGinley 2025). 

Virginia is one of a handful of Southern states that for decades explicitly banned public employees and employers from entering into collective bargaining agreements. In 2020, Virginia took an important step toward making collective bargaining newly optional for local governments, but the state’s policies remain out of step compared with most states. Virginia lacks a statewide collective bargaining statute covering all local government employees and still has a ban in place barring state employees from collective bargaining (Borja 2022).

As shown in Table 1, the majority of states and Washington, D.C., already ensure collective bargaining rights for most public employees, including state workers. Many states have had statewide public-sector bargaining statutes in place for decades (Rueben 1996; Sanes and Schmitt 2014). Such laws typically set clear, uniform guidelines for union elections and contract negotiation processes and establish state labor boards charged with fostering productive labor-management relations (including timely contract settlements), ensuring broad awareness of and compliance with statutory guidelines, and mediating or adjudicating disputes as needed.

Table 1

Who are public employees in Virginia?

In 2024, approximately 560,000 Virginians worked in state and local government occupations (BLS-CES 2020–2025). This includes teachers and school staff, firefighters, transit operators, law enforcement, administrative staff, and employees serving the state’s public safety, transportation, health care, judicial, corrections, and higher education systems. As shown in Table 2, 21.2% of Virginia state government employees and 20.4% of Virginia local government employees are Black, and 57.3% of state government employees and 64.5% of local government employees are women. Many public-sector workers in the state are highly educated. Virginia public-sector workers are more than twice as likely to have advanced degrees as private-sector workers and are much less likely to have a high-school level or less education.

Table 2

Anti-union state policies are rooted in racism and harm all Virginia workers

Virginia’s ban on union contracts for public employees is a Jim Crow-era policy with deep roots in the history of slavery and white supremacy. After the passage of federal labor laws accelerated worker organizing in the 1930s, Virginia joined several Southern states in adopting anti-union state laws designed to prevent multiracial union organizing and suppress Black workers’ wages and power (Childers 2023). The Virginia Assembly first took an explicit stance on public-sector bargaining in response to the unionization of Black hospital employees at the University of Virginia in 1946, via a joint resolution declaring it against the public policy of the state to negotiate with public employee unions (a stance later affirmed by state supreme court decisions and codified in statute in 1993). Virginia was also among the first states to adopt anti-union so-called right-to-work legislation in 1947, an anti-labor policy jointly promoted by white supremacist organizations and industry groups intent on slowing the growth of unions to maintain access to cheap labor—especially in Southern states (The Commonwealth Institute 2022; Watts 2021; Pierce 2017, 2018; Sherer and Gould 2024).

As a result of these long-standing anti-union state policies, unionization rates in Virginia for both public- and private-sector workers are well below national averages.

Figure A

Narrowing Virginia’s large public-sector pay gap

Across the country, public-sector employees earn less than their private-sector counterparts, and this pay gap has widened in recent years. In the latest available five-year period (September 2020–August 2025), state and local government employees earned, on average, 17.2% less than private-sector employees with similar education and experience. The size of the public-sector pay gap varies across states and is largest in states like Virginia where most public employees lack collective bargaining rights. Public-sector workers with strong bargaining rights experience a narrower pay gap (-14.3%) than those with weak (-19.6%) or no bargaining rights (-22.5%) (see Table 3).

Virginia currently has one of the largest public-sector pay gaps in the nation. Among all 50 states, Virginia’s -26.7% public-sector pay gap appears to be the second highest, though differences among states clustered at the bottom of the rankings are not statistically significant. (Pay gap statistics are for full-time wage and salary workers ages 18–64, based on the authors’ analysis of pooled September 2020–August 2025 Current Population Survey microdata downloaded from Flood et al. 2025 and EPI 2025a.)

Compensation packages of public employees, on average, include more robust benefits than those of private-sector workers, but Virginia’s public-sector compensation gap remains large, even when factoring in more robust benefits. Benefits are an estimated 32.0% of pay for private-sector workers in the South Atlantic region and an estimated 51.3% of pay for Virginia public-sector workers. Factoring in benefits, Virginia’s public-sector compensation gap shrinks to a still sizable 16.0% (authors’ estimate based on BLS-ECEC 2020–2025 and Public Plans Data 2020–2024; see Morrissey and Sherer 2024 for methodology).1

Table 3

Addressing high staff vacancy and turnover rates in Virginia’s public sector

The growing public-sector pay gap has become a particular concern at a moment when low pay has created serious challenges to recruiting and retaining teachers and other public employees across the country (Cooper and Martinez Hickey 2022; MissionSquare Research Institute 2023; Wething 2024a, 2024b; Martinez Hickey 2025). Virginia has faced particularly acute staffing shortages in public education and in units of state government in recent years (Manzanares 2025; Cantor 2025).

The latest (2025) biennial compensation report from the state’s Department of Human Resource Management notes that “Years without any salary adjustments in the past have made it difficult for state agencies to build a proactive and sustainable approach to addressing compensation, recruitment and retention concerns” (VDHRM 2025). Between 2001 and 2023, low average salary increases for state workers (just 2.9% per year, compared with 3.4% annually in the private sector) have led to high vacancy and turnover rates. As of 2024, roughly 1 in 5 (22.4%) state jobs were unfilled, and the median salary across the state workforce was just $61,305—a full $5,000 less than the median city employee salary in Richmond, which has adopted its own collective bargaining ordinance (McGinley 2025).

Chronic state employee staffing shortages are already having direct impacts on the health, safety, and quality of life of Virginians. To take one example, in 2025 the Virginia Department of Juvenile Justice (DJJ) drew headlines after reporting that inability to adequately staff large facilities for incarcerated youth had led to unsafe conditions, lockdowns, increased restrictions on out-of-cell time, and a lack of rehabilitative services. DJJ directors indicated ongoing difficulty recruiting for open positions, despite participating in job fairs, college and university visits, outreach to military and veteran communities, and offering signing bonuses and referral incentives (Manzanares 2025). The root causes of understaffing identified by Virginia state agencies like DJJ—from burnout and high workloads to low starting salaries, lackluster raises, and difficult or unsafe working conditions—are precisely the topics that a structured collective bargaining process would allow state agencies to address, with direct input from frontline employees.

Addressing racial and gender pay gaps to improve recruitment and retention of workers of color and women

The public-sector pay gap disproportionately affects Black workers and women, who are more likely to be employed in public-sector jobs and who are disadvantaged in the broader labor market (Childers 2025). Strengthening collective bargaining rights for government workers in Virginia therefore promises to narrow the pay gap and reduce racial and gender inequalities in public institutions and across the labor market.

For example, a 2024 RAND report showed that Black teachers nationally receive lower average salaries and pay raises than white teachers do, a difference linked directly to the fact that Black teachers were less likely to live in states where public educators had collective bargaining rights. The inadequacy of pay is one of the main reasons teachers report for leaving the profession, further contributing to the demographic mismatch between teachers and students (e.g., nationwide over half of all students are children of color, but the teaching workforce remains around 80% white) (Steiner et al. 2024; Gopalan 2025). Likewise, data show that in states like Wisconsin where legislators have weakened formerly strong collective bargaining rights in the past two decades, resulting decreases in unionization levels and worker wages have measurably widened public-sector pay gaps and gender pay gaps (Nack et al. 2019; García and Han 2021; Biasi and Sarsons 2022).

Building on success and remedying limitations of current state law that only permits local collective bargaining

In 2020, Virginia partially lifted its long-standing ban on public-sector collective bargaining with legislation creating an “opt-in” system that has allowed local governments to set their own policies on whether and how to bargain with their own employees. While this opening fell far short of creating a consistent statewide framework for collective bargaining, it has resulted in at least 17 of Virginia’s largest cities, counties, and school boards adopting collective bargaining ordinances and creating new pathways to union contracts for substantial numbers of public employees—including, for example, 27,000 teachers and school staff and 12,000 county employees in Fairfax County; nearly 4,000 City of Richmond employees; and others (Borja 2022; Sharma 2022; Khalil 2023; Lukert 2024; Pope 2025; Walter 2026).

The local “opt-in” system was a positive step forward that has already revealed high levels of interest in collective bargaining among Virginia workers. The system resulted in new union contracts covering tens of thousands of frontline educators and civil servants, providing an initial boost to Virginia’s historically low unionization rate.

Significant limitations of the new “opt-in” system have also quickly become clear. As noted above, the major shortcoming of the current law is that it does not ensure equal collective bargaining rights for all public employees. Virginia state employees remain barred from collective bargaining, and in local government where collective bargaining is permitted (but not required), workers continue to lack collective bargaining rights, unless they are able to persuade local officials to adopt and then implement a collective bargaining ordinance. When localities do adopt such ordinances, they may vary in strength and effectiveness (Overman 2023).

As a result, Virginia’s current “opt-in” system for local collective bargaining has generated an uneven patchwork of highly variable (and potentially unstable) collective bargaining policies across the state. Some local governments have continued to block workers’ path to a union contract by rejecting appeals from their own employees to adopt local collective bargaining ordinances (Murphy 2024; Cooper 2025; Lytle 2025; Wilkinson 2025). Because current state law leaves the burden of collective bargaining policy development up to each individual local government, some jurisdictions have expressed interest in or support for collective bargaining while remaining reluctant to invest the necessary time or scarce administrative or legal resources to developing and implementing a local ordinance. Even in larger local jurisdictions with strong collective bargaining ordinances now in place, the “opt-in” system remains fragile and highly vulnerable to instability whenever turnover occurs among elected leaders or administrators with experience necessary to maintain unique local labor-management systems.

Creating a state labor board to provide efficiency and stability for all Virginia public employers and employees

A proposed state labor board equipped to administer a statewide, uniform collective bargaining framework would serve all Virginia state and local government entities and provide consistency, efficiency, stability, and economies of scale. All Virginia public employers and employees would benefit from access to a central, independent state board with capacities to advise public employers and employees about collective bargaining procedures, administer union elections, and mediate contract negotiations. The creation of such state capacities is also especially important at moment when federal labor agencies with similar capacities have been eliminated or rendered non-functional. Indeed, across the country, many states are relying more than ever on their existing state labor boards, and in some cases, are exploring paths to expanding state labor board capacities in order to ensure consistent protections of workers’ rights to unionize and collectively bargain (H.B. 3005; Walter and Madland 2025).

Conclusion

In 2026, Virginia lawmakers should seize the opportunity to enact strong, comprehensive collective bargaining legislation that covers all state and local government workers and creates a state labor board to administer the new system. Under Virginia’s current state law (where collective bargaining is banned for state employees and allowed only for some local government workers), pay for Virginia public employees has lagged far behind that of private sector counterparts with similar education and experience.

Stronger collective bargaining rights can help shrink Virginia’s large public-sector pay gap, reduce racial and gender pay gaps, and improve recruitment and retention of qualified public employees. Removing barriers to unionization for public employees is also a critical step toward reversing the impacts of long-standing anti-worker state policies in Virginia that have for decades suppressed all workers’ wages and contributed to growing income inequality. Lastly, state action to shore up public employee rights is especially important at moment when the federal government is attacking civil servants, public education, health care, and all public services. By extending full collective bargaining rights to historically excluded state and local government workers, state lawmakers can help lead the way to a more vibrant, equitable economy rooted in multiracial democracy in Virginia, the South, and the nation.

Notes

1. If anything, our comparison likely minimizes the public-sector compensation gap in Virginia by comparing benefits for private-sector workers in the South Atlantic region to public-sector workers throughout the country because public-sector data for the South Atlantic region is not available from the Bureau of Labor Statistics. Our comparison does account for the fact that Virginia pension benefits are less generous than public pensions in many parts of the country, though possibly not in two Northern Virginia counties with their own retirement systems. It does not fully account for other differences in benefits between Virginia public-sector workers and their counterparts in other states, though it does account for the fact that public-sector workers in Virginia are covered by Social Security (not true in some states) and adds 1% for public-sector retiree health benefits that are not included in BLS compensation statistics. Finally, it compares all public-sector workers with all private-sector workers, when arguably the better comparison would be between public-sector workers and private-sector workers employed by large employers, who tend to provide more generous benefits than small employers.

References

Biasi, Barbara, and Heather Sarsons. 2022. “Flexible Wages, Bargaining, and the Gender Gap.” Quarterly Journal of Economics 137, no. 1 (February): 215–266. https://doi.org/10.1093/qje/qjab026.

Bivens, Josh, and Heidi Shierholz. 2018. What Labor Market Changes Have Generated Inequality and Wage Suppression? Employer Power Is Significant but Largely Constant, Whereas Workers’ Power Has Been Eroded by Policy Actions. Economic Policy Institute, December 2018.

Borja, Mel. 2022. “How Public-Sector Workers Are Building Power in Virginia.” Working Economics Blog (Economic Policy Institute), February 18, 2022.

Brannick, Priya. 2019. The Janus Impact and Grading of State Public Sector Labor Laws. Commonwealth Foundation, August 2019.

Brannick, Priya M., and Andrew Holman. 2022. The Battle for Worker Freedom in the States: Grading State Public Sector Labor Laws. Commonwealth Foundation, September 2022.

Bureau of Labor Statistics (BLS-CES). 2020–2025. Current Employment Survey, State and Area Employment Data. Accessed January 2026.

Bureau of Labor Statistics (BLS-ECEC). 2020–2024. Employer Costs for Employee Compensation data. Accessed January 2026.

Cantor, Dave. 2025. “Virginia Corrections Department has 2,400 Open Positions.” Virginia Public Radio, August 14, 2025.

Childers, Chandra. 2023. Rooted in Racism and Economic Exploitation: The Failed Southern Economic Development Model. Economic Policy Institute, October 2023.

Childers, Chandra. 2025. “Disinvestment in the Public Sector Undermines Opportunities for Black Women Across the South: Trump Cuts Further Threaten Key Services for Working People Across the Nation.” Working Economics Blog (Economic Policy Institute), July 7, 2025.

Commonwealth Foundation. 2021. “Public Sector Labor Law Database.” Accessed May 9, 2021.

Cooper, David, and Sebastian Martinez Hickey. 2022. Raising Pay in Public K–12 Schools Is Critical to Solving Staffing Shortages: Federal Relief Funds Can Provide a Down Payment on Long-Needed Investments in the Education Workforce. Economic Policy Institute, February 2022.

Cooper, Tracy. 2025. “Portsmouth Workers Frustrated by Private Collective Bargaining Talks.” 13 News Now, July 22, 2025.

Dippel, Christian, and Zachary Sauers. 2019. “Does Increased Union Power Cause Pension Under-Funding in the Public Sector?” (preliminary draft). September 23, 2019.

Economic Policy Institute. 2025a. Current Population Survey Extracts, Version 2025.12.18, https://microdata.epi.org.

Economic Policy Institute. 2025b. How Many Federal Employees Live in Your State? Accessed January 12, 2026.

Flood, Sarah, Miriam King, Renae Rodgers, Steven Ruggles, J. Robert Warren, Daniel Backman, Etienne Breton, Grace Cooper, Julia A. Rivera Drew, Stephanie Richards, David Van Riper, and Kari C.W. Williams. IPUMS CPS: Version 13.0 . Minneapolis, MN: IPUMS, 2025. https://doi.org/10.18128/D030.V13.0.

Frandsen, Brigham R., and Michael Webb. 2017. “Public Employee Pensions and Collective Bargaining Rights: Evidence from State and Local Government Finances.” Hutchins Center Working Paper no. 35, The Brookings Institution, October 2017.

García, Emma, and Eunice Han. 2021. The Impact of Changes in Public-Sector Bargaining Laws on Districts’ Spending on Teacher Compensation. Economic Policy Institute, April 2021.

Gopalan, Maithreyi. 2025. “Why Does Teacher-Student Demographic Representation Matter for Student Success?” HEDCO Institute for Evidence-Based Educational Practice, October 15, 2025.

Han, Eunice S. 2019. “The Impact of Teachers Unions on Teachers’ Well-Being Under Various Legal Institutions: Evidence from District–Teacher Matched Data.” American Educational Research Association 5, no. 3 (August). https://doi.org/10.1177/2332858419867291.

H.B. 1263, 2026 H.D., Reg. Sess. (Va. 2026).

H.B. 2764, 2025 H. D., Reg. Sess. (Va. 2025).

H.B. 3005, 2025 Gen. Assemb. (Ill. 2025).

International Association of Fire Fighters (IAFF). n.d. “Collective Bargaining” (web page). Accessed May 2024.

Khalil, Jahd. 2023. “Hundreds of Richmond City Employees Select Unions.” Virginia Public Media, June 23, 2023.

Lukert, Luke. 2024. “Fairfax Co. Teachers Vote Overwhelmingly for Collective Bargaining Rights.” WTOP, June 11, 2024.

Lytle, Derek. 2025. “Virginia Beach School Board Sits Divided on Collective Bargaining.” 13 News Now, November 11, 2025.

Manzanares, Keyris. 2025. “Virginia Juvenile Justice System Strained by Staffing Shortages.” Virginia Public Media, April 9, 2025.

Martinez Hickey, Sebastian. 2025. “The School Bus Driver Shortage Has Improved Slightly but Continues to Stress K–12 Public Education.” Working Economics Blog (Economic Policy Institute), November 3, 2025.

McGinley, Sophie. 2025. Building a More Equitable Commonwealth: The Case for Collective Bargaining Rights for Virginia State Employees. The Commonwealth Institute, August 2025.

McNicholas, Celine, Margaret Poydock, Heidi Shierholz, and Hilary Wething. 2025. Unions Aren’t Just Good for Workers—They Also Benefit Communities and Democracy. Economic Policy Institute, August 2025.

McNicholas, Celine, Lynn Rhinehart, Margaret Poydock, Heidi Shierholz, and Daniel Perez. 2020. Why Unions Are Good for Workers—Especially in a Crisis Like COVID-19: 12 Policies That Would Boost Worker Rights, Safety, and Wages. Economic Policy Institute, August 2020.

Mishel, Lawrence, and Josh Bivens. Identifying the Policy Levers Generating Wage Suppression and Wage Inequality. Economic Policy Institute, May 2021.

MissionSquare Research Institute. 2023. State and Local Government Employment Trends, 2023 (issue brief). May 17, 2023.

Morrissey, Monique, and Jennifer Sherer. 2024. The Public-Sector Pay Gap Is Widening. Unions Help Shrink It. Economic Policy Institute, August 2024.

Murphy, Ryan. 2024. “Virginia Beach Rejects Collective Bargaining for City Employees.” WHRO Public Media, April 30, 2024.

Nack, David, Michael Childers, Alexia Kulwiec, and Armando Ibarra. 2019. “The Recent Evolution of Wisconsin Public Worker Unionism Since Act 10.” Labor Studies Journal 45, no. 2: 147–165. https://doi.org/10.1177/0160449X19860585.

National Conference of State Legislators (NCSL). n.d. “Collective Bargaining database.” Accessed May 2024.

National Council on Teacher Quality (NCTQ). 2019. “Collective Bargaining Laws database.” Updated January 2019.

National Education Association (NEA). 2020. “Collective Bargaining Laws for Public Sector Education Employees,” November 2020 (unpublished document used by permission).

New Mexico Public Employee Labor Relations Board. 2023. Public Sector Collective Bargaining by State.

Oakford, Patrick, and Margaret Poydock. 2025. “Trump Is the Biggest Union-Buster in U.S. History: More Than 1 Million Federal Workers’ Collective Bargaining Rights Are at Risk.” Working Economics Blog (Economic Policy Institute), September 2, 2025.

Overman, Stephenie. 2023. “In Virginia, ‘Patchwork’ of Ordinances Makes Public-Sector Organizing a Maze.” Virginia Mercury, January 16, 2023

Pierce, Michael. 2017. “The Origins of Right-to-Work: Vance Muse, Anti-Semitism, and the Maintenance of Jim Crow Labor Relations.” The Labor and Working-Class History Association, January 12, 2017.

Pierce, Michael. 2018. “Vance Muse and the Racist Origins of Right-to-Work.” Expert Forum (American Constitution Society), February 22, 2018.

Pope, Michael. 2025. “Fairfax County May Soon Be the Home of the Largest Collective Bargaining Agreement in Virginia.” WVTF, September 10, 2025.

Public Plans Data. 2001–2024. Center for Retirement Research at Boston College, MissionSquare Research Institute, National Association of State Retirement Administrators, and the Government Finance Officers Association. Accessed January 5, 2026.

Rueben, Kim. 1996. “Extended NBER Public Sector Collective Bargaining Law Data Set” [Stata and Excel files]. Downloadable data available at www.nber.org/research/data/nber-public-sector-collective-bargaining-law-data-set (see last paragraph on this web page). Accessed January 12, 2026.

Ruggles, Steven, Sarah Flood, Matthew Sobek, Daniel Backman, Grace Cooper, Julia A. Rivera Drew, Stephanie Richards, Renae Rodgers, Jonathan Schroeder, and Kari C.W. Williams. IPUMS USA: Version 16.0 . Minneapolis, MN: IPUMS, 2025. https://doi.org/10.18128/D010.V16.0

Sanes, Milla, and John Schmitt. 2014. Regulation of Public Sector Collective Bargaining in the States. Center for Economic and Policy Research, March 2014.

S.B. 378, 2026 S., Reg. Sess. (Va. 2026).

S.B. 917, 2025 S., Reg. Sess. (Va. 2025).

Sharma, Rahul Chowdry. 2022. “Where Can Public Sector Employees Collectively Bargain in Virginia? Increasingly, the Commonwealth’s Most Populous Cities and Counties.” Virginia Mercury, July 28, 2022.

Sherer, Jennifer, and Elise Gould. 2024. “Data Show Anti-Union ‘Right-to-Work’ Laws Damage State Economies: As Michigan’s Repeal Takes Effect, New Hampshire Should Continue to Reject ‘Right-to-Work’ Legislation.” Working Economics Blog (Economic Policy Institute), February 13, 2024.

Shimabukuro, Jon O., and Julie M. Whittaker. 2014. Federal Labor Relations Statutes: An Overview. Congressional Research Service (CRS) R42526. Updated September 5, 2014.

Steiner, Elizabeth D., Ashley Woo, and Sy Doan. 2024. Larger Pay Increases and Adequate Benefits Could Improve Teacher Retention: Findings from the 2024 State of the American Teacher Survey. RAND Institute, November 20, 2024.

The Commonwealth Institute. 2022. History of Labor in Virginia: An Interactive Timeline and Map. Accessed January 12, 2026.

Valletta, Robert G., and Richard B. Freeman. 1988. “The NBER Public Sector Collective Bargaining Law Data Set.” Appendix B in When Public Employees Unionize, edited by Richard B. Freeman and Casey Ichniowski. NBER and Univ. of Chicago Press.

Virginia Department of Human Resource Management (VDHRM). 2025. Biennial Compensation Report. November 2025.

Walter, Karla. 2026. Virginia Workers’ Biggest Win in Decades Could Come in 2026. Center for American Progress, January 12, 2026.

Walter, Karla, and David Madland. 2025. Union Trigger Laws 101 How States Can Protect Workers if Federal Labor Law Falls. Center for American Progress, November 19, 2025.

Watts, Parker. 2021. “Labor Day Reflections on Race, Power, and Organized Labor in Virginia.” The Commonwealth Institute, September 1, 2021.

Wething, Hilary. 2024a. “Today’s Teacher Shortage Is Just the Tip of the Iceberg: Part I.”“Today’s Teacher Shortage Is Just the Tip of the Iceberg: Part I.” Working Economics Blog (Economic Policy Institute), October 9, 2024.

Wething, Hilary. 2024b. “Today’s Teacher Shortage Is Just the Tip of the Iceberg: Part II.” Working Economics Blog (Economic Policy Institute), October 16, 2024.

Wilkinson, Nolan. 2025. “Proposal to Allow Frederick City Employees to Unionize Tabled.” The Frederick News-Post, September 25, 2025.