Key takeaways

• The label “right to work” is intentionally misleading. RTW laws provide no job protections. What they do is make it harder for workers to form and sustain strong unions.
• RTW laws are intended to weaken unions. They are linked to lower unionization rates, lower wages and benefits, and worse safety outcomes for all workers.
• RTW laws have racist, anti-union origins dating to the 1940s. For decades, big business and white supremacist groups promoted RTW laws to diminish workers’ collective bargaining rights and block multiracial organizing.
• RTW laws don’t boost job growth but do erode job quality and increase inequality.
• Trends on RTW are reversing. No state has adopted a new RTW law since 2017, and multiple states have repealed or rejected RTW.

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.

Figure A

These increases continue to be crucial for low-wage workers as they contend with the affordability crisis. The average increase in annual wages for a full-time, year-round worker resulting from these minimum-wage hikes ranges from $573 in Oregon to $811 in Alaska. These pay raises directly boost workers’ incomes, giving them a leg-up in the race against rising prices—a straightforward example of how policymakers can often more easily tackle affordability challenges through policy decisions that boost wages, such as setting strong wage floors.

A higher minimum wage has a positive impact on more than just the workers who currently earn the minimum wage; indirectly affected workers will see their pay go up too as employers adjust their wage ladders to the new wage floor. Our analysis of the increases in Alaska, Oregon, and Washington, D.C., accounts for these “spillover” effects and finds:

• Women make up more than half (56.3%) of affected workers.
• The wage increases disproportionately benefit Black and Hispanic workers. Black workers make up 15.3% of affected workers, despite making up 10.4% of the workforce across the three areas. Hispanic workers make up a similar percentage of the workforce (12.4%) but make up more than a fourth (26.0%) of affected workers.
• The vast majority (89.3%) of affected workers are age 20 or older, and more than 3 in 5 workers (62.1%) are 25 or older.
• More than half (52.9%) of the affected workers work full-time.
• The increases will raise wages for those who need it the most. Half (50.5%) of affected workers belong to households whose incomes are less than 200% of the poverty line.
• More than 1 in 5 (23.6%) of affected workers are parents.

Table 1

State and local policymakers should combat cost-of-living challenges with minimum wages

Because costs of living can vary significantly within a state, local policymakers should establish strong wage floors if the state minimum wage is inadequate for their area. In June, city councilors in Albuquerque, New Mexico, took an important step in that direction, passing a minimum-wage increase to $15 an hour by 2029. Whereas the current state minimum in New Mexico is $12 an hour, EPI’s Family Budget Calculator shows that a living wage for a single adult working full-time in Albuquerque is $17.56 an hour.1 The proactive step taken by local elected officials will provide significantly more economic security for low-wage workers in the city.

This summer, the minimum wage will increase in localities in California, Illinois, and Maryland (Table 2) due to “indexing”—automatic annual adjustments written into the minimum-wage law that require the wage floor be adjusted for price increases each year. Without indexing, the minimum wage is worth less and less every year as prices rise. With time, this can dramatically erode the value of the minimum wage. For example, the federal minimum wage of $7.25 an hour—which has not increased since 2009—has now lost 30% of its purchasing power.

Table 2

Indexing to inflation is a commonsense policy enacted in dozens of cities and states across the country, but it is not the strongest way to protect the value of the minimum wage. Adjusting for price increases mostly protects the real value of a low-wage worker’s paycheck (although low-wage workers are more vulnerable to inflation than top-of-the-line inflation measures indicate). However, in a well-functioning economy, wages for most workers—especially higher wage workers—will grow faster than inflation. If the minimum wage only rises at the rate of price growth, then over time, this can increase inequality between low-wage workers and everyone else.

One way to address this issue and ensure that low-wage workers do not fall further away from the middle class is to index the minimum wage to median wage growth. EPI’s latest vision for a federal minimum wage explicitly targets a wage floor that rises to two-thirds of the median wage and remains indexed there thereafter. This policy both creates a high floor that provides significantly higher wages for workers across the country and protects the minimum wage’s value against price increases and increases in wage inequality over time.

Oklahoma fails to pass minimum-wage ballot measure

In June, a minority of eligible Oklahoma voters rejected a ballot initiative (State Question 832) that would have increased the state’s minimum wage to $15 an hour by 2029. The initiative’s failure is a costly missed opportunity to increase wages for Oklahoma workers. More than 1 in 5 workers in the state earn less than $15 an hour, and if passed, the ballot initiative would have provided more than $783 million in increased earnings for low-wage workers.

During a time when cost of living and inflation are some of the most important concerns for voters, Oklahomans might have worried that increasing the minimum wage would have hurt affordability in the state. In fact, the opposite is true. Although raising the minimum wage can lead some affected businesses to increase prices, the resulting price increases are extremely modest—far smaller than the increase in pay that would go to low-wage workers. Even some of the most ambitious wage floor policies, such as California’s $20 fast food minimum wage, only increased fast-food prices 2.1% (around eight cents for a $4 item). A 10% increase in the minimum wage is associated with a 0.14 percentage point increase in CPI increase. In contrast, a 10% minimum-wage increase boosts income at the 10th percentile by around 3.6%, an order of magnitude greater. The average full-time, year-round Oklahoman worker affected by SQ 832 would have gained $2,322 in annual wages if voters had approved the initiative.

It is also important to highlight SQ 832’s winding path to the ballot. Ballot initiatives have historically been an important mechanism for passing minimum-wage increases in states with conservative-dominated legislatures like Florida, Missouri, and Nebraska. While Oklahoma is one of only three states in the South that has a ballot initiative process, conservative politicians have been increasingly curtailing it.

The Oklahoma minimum-wage ballot initiative is a prime example of this. Advocates originally began collecting signatures for SQ 832 in the lead-up to the November 2024 general election. Despite collecting nearly twice the necessary signatures in enough time to qualify for the ballot, advocates were thwarted when Governor Kevin Stitt delayed the State Question until the June 2026 gubernatorial primary election, a low turnout election in contrast to a general election with presidential candidates on the ballot. Voter turnout in the 2026 primary election was 26%, around half of what it was in the 2024 election.

In response to the emergence of SQ 832, the Oklahoma legislature also passed restrictions on the signature-gathering process, limiting the number of signatures that can be gathered from populous areas like Tulsa and Oklahoma City. These restrictions will make it more costly and logistically challenging to pass a future minimum-wage increase in Oklahoma. With such low voter turnout and marked interference in the ballot initiative process, it is difficult to say that SQ 832’s failure reflects a lack of popular support for minimum-wage increases in Oklahoma. Regardless of the cause, without a future minimum-wage increase, the issue of low pay in Oklahoma is only going to grow.

When policymakers like those in Oklahoma fail to adequately set the wage floor, it suppresses worker pay, not just for the lowest-paid workers, but for low-wage workers in general. Workers need a raise to help them overcome the affordability crisis, and the minimum wage is an evidence-backed tool under policymakers’ control to help them do that.

1. Assuming 81% of income comes from wages.

Does raising the minimum wage increase unemployment?

In brief: No. High-quality economic research finds increasing the minimum wage does not significantly impact employment.

In detail: The 90% of high-quality economic studies show that increasing the minimum wage boosts wages for low-wage workers without meaningfully increasing unemployment. These studies use statistical tools and empirical methods to measure what happens to workers before and after a minimum wage increase, controlling for other factors that can impact employment. The consistency of these findings across time, place, and level of increase is powerful evidence that increasing the minimum wage creates a healthier low-wage labor market.

An increase in the minimum wage raises the cost of labor for businesses by definition, but the economy can absorb these changes through channels of adjustment including decreased turnover, modest price increases (see Question 2), lower profits, and the reallocation of workers to more productive firms. Even if a minimum wage increase leads businesses to adjust their staffing levels, what workers are likely to experience are decreases in hours worked or increased time between jobs, not categorical unemployment. Higher hourly earnings can more than offset these reductions, leaving many workers with greater total income even if they are working fewer hours.

Will raising the minimum wage cause inflation?

In brief: No. Increasing the minimum wage does not meaningfully increase prices.

In detail: Economists do find that raising the minimum wage increases prices at affected businesses but only very modestly. For example, one study found that a 10% minimum wage increase was associated with a 0.14 percentage point increase in the Consumer Price Index. A study focused on the restaurant industry found a 10% increase in the minimum wage was associated with a 0.58% menu price increase. Even some of the most ambitious minimum wage policies, such as California’s $20 hourly wage floor for fast-food workers, only increased fast-food prices 2.1% (around 8 cents for a $4 item). 

The economic benefits of the minimum wage far exceed these price increases. For low-wage workers, the wage boost from the higher minimum wage more than compensates for increased prices of the goods and services they buy. These workers are in turn spending more in aggregate because of their additional income, which can boost the overall economy. It is also worth noting that modest price increases on things like restaurant menu items are redistributive. Higher-earning consumers pay higher prices, transferring income to low-wage workers receiving bigger paychecks.

Claims that increasing the minimum wage will dramatically increase prices in the economy are false. Low-wage labor is a small share of total business expenses. In restaurants, for example, total labor costs are around 30% of operating costs—which also include rent, food, utilities, and insurance—and the wage bill of the lowest paid workers is an even smaller amount. This, combined with the fact that minimum wage increases can be offset through reduced profits, lower turnover, and higher productivity, is why price increases are small.

Will businesses just relocate if a state or locality raises its minimum wage?

In brief: The best economic research suggests businesses do not move in response to minimum wage increases.

In detail: One way that economists try to understand the impact of minimum wage increases is to compare economic outcomes across jurisdictional borders where a minimum wage increase took effect on one side but not the other. An analysis of cross-state and county impacts of all local minimum wage differences between 1990 and 2006 found no evidence that employment decreased in the places where the minimum wage went up or that employment increased in the places without a minimum wage increase. More recent research supports these findings, strongly suggesting that businesses are not relocating or moving their workers in response to minimum wage changes.

Businesses commonly affected by minimum wage changes (such as restaurants and retail) want to locate where there are consumers with money to spend. Because raising the minimum wage boosts the spending power of low-income households, it can strengthen the local customer base for these direct-to-consumer businesses even as it raises their labor costs.

Is the minimum wage an effective way to fight poverty?

In brief: Increasing the minimum wage does reduce poverty and should be paired with a strong safety net.

In detail: Minimum wage increases do significantly reduce poverty by boosting household income, especially among low-income households. Research has found that a 10% increase in the minimum wage reduces nonelderly poverty by 2–4%. When EPI applied this research to the 2021 Raise the Wage Act (which would have gradually increased the minimum wage to $15 an hour), an estimated 1.8 to 3.7 million individuals would have been lifted out of poverty, including up to 1.3 million children.

The current weakness of the federal wage floor exposes workers to poverty-level wages. As of 2025, a full-time worker earning the federal minimum wage makes less than the poverty line. A stronger wage floor would generate more savings across critical safety net programs like Medicaid, SNAP, the Earned Income Tax Credit (EITC), and the Child Tax Credit (CTC), as fewer workers would need or be eligible for these programs due to their increased wages. The safety net would thus be more targeted toward the households that need assistance the most. Those public savings can and should be reinvested in those programs to make benefits more generous.

Can increasing the minimum wage harm workers by pushing them over a “benefits cliff”? 

In brief: The minimum wage does reduce safety net program eligibility, but the wage gains almost always outweigh the loss of benefits.

In detail: Most income-tested safety net programs are not characterized by “cliffs” but rather gradual phase outs. Nevertheless, when a worker’s income increases because of a minimum wage increase, they can lose eligibility to programs like Medicaid, EITC, CTC, SNAP, and housing assistance. Research on the interaction between the minimum wage and safety net eligibility finds that benefit reduction is significantly outweighed by the increase in income from the minimum wage. Benefit reductions offset around a third of the income increases for low-income families caused by the minimum wage—meaning that on net they still benefit overwhelmingly.

Higher minimum wages also help low-wage workers increase their access to medical coverage. While minimum wage increases can lift workers out of income eligibility for Medicaid, they simultaneously increase the take-up of employer-based health insurance plans by many low-wage workers, since those workers can more easily afford those plans. Many workers who lose Medicaid eligibility also can receive heavily subsidized private health care through the Affordable Care Act (ACA) exchanges. Researchers estimate that the $20 fast-food minimum wage in California could push almost 60% of Medi-Cal eligible workers in the industry off Medi-Cal and onto alternative health insurance. The wage benefits of the wage floor increase far outweigh the annual premium contributions workers must pay for ACA marketplace healthcare, even accounting for the Trump administration’s choice to let expanded subsidies for the ACA expire.

The reduction in public benefit usage created by higher minimum wages generates substantial savings for federal and state government. These savings should be reinvested in the safety net by expanding program eligibility or otherwise strengthening programs.

There are some safety net programs, which have cliff-like characteristics, like child care assistance, although states are required to provide a graduated phase-out. In rare cases where changes in program eligibility from a minimum wage increase does reduce a family’s total income, this indicates that the design of these programs’ eligibility criteria needs reform, not that the minimum wage increase should be abandoned.

Should lower cost of living in the South and Midwest mean a lower wage floor in those states?

In brief: Even accounting for differences in the cost of living, the minimum wage is far too low in many states across the South and Midwest.

In detail: Cost of living does vary between states and regions across the country, but the minimum wage is still too low across the South and the Midwest. According to EPI’s Family Budget Calculator, even under conservative assumptions of what constitutes a living wage, there is almost no county in the U.S. where a single adult worker can achieve a modest but adequate standard of living earning less than $15 an hour. In fact, many metro areas in the South and Midwest are much more expensive to live in. The living wage in Austin, Atlanta, and Charlotte exceeds $20 an hour. Affordability is still a pressing issue across these regions, even if on average the cost of living is lower. Notably, voters in Florida, Missouri, and Nebraska passed ballot referenda to raise their state minimum wages to $15.

Low wage floors in Southern and Midwestern states hurt workers by suppressing their pay. Most of the states that use the $7.25 federal minimum wage are in the South and Midwest, meaning the effective wage floor in these states is a poverty-level wage (see Question 5). Compounding the problem, many states in these regions preempt localities from passing their own minimum wage policies, preventing policymakers in these jurisdictions from setting wage floors that meet the needs of workers. The use of preemption to dismantle higher labor standards like the minimum wage in the South and Midwest has a long history of being used to reinforce anti-Black racism and white supremacy in these regions.

Can employers ever pay less than the minimum wage?

In brief: Most U.S. minimum wage laws do exempt some groups of workers (such as farmworkers) or set lower minimum wages that apply in certain circumstances (such as for workers who customarily receive tips). Unfortunately, these exemptions can be deeply harmful to workers; in some cases, they were originally adopted to exclude workers of color from minimum wage protections.

In detail: Federal and state labor standards make several groups of workers either ineligible for minimum wage protections or subject to a separate “subminimum wage.”

The FLSA exempts a variety of occupations and types of workers from minimum wage protections. Agricultural workers are excluded from the federal minimum wage entirely and workers who customarily receive tips may be paid a subminimum wage (sometimes called the “tipped minimum wage”) as low as $2.13 an hour (see Questions 8–11). Both of these exemptions originated as ways to exclude Black workers from New Deal economic policies in order to appease Southern lawmakers. Originally, the FLSA also excluded domestic workers, another group of workers with a high concentration of Black workers, but lawmakers extended coverage to them in 1974.

The FLSA also allows employers who have been granted a certificate from the U.S. Department of Labor to pay less than the minimum wage to employees with disabilities (see Question 14).

Another category of federal exemptions and subminimum wages impact young workers. Youth under 20 can be paid as little as $4.25 per hour for their first 90 calendar days of employment. Full-time students, apprentices, and student-learners can also be subject to subminimum wages. And specific occupations typically held by young workers, like babysitters and seasonal amusement workers, are exempt.

Workers misclassified as independent contractors, such as gig economy workers and other wrongly classified employees are not eligible for FLSA protections, including the minimum wage. Also, despite the fact around half of incarcerated people work full-time, these individuals are also excluded from the minimum wage.

State and local policymakers in many states have made efforts to close many of these gaps in minimum wage coverage, but states that do not go beyond the federal standards maintain these exemptions.

Does raising the tipped subminimum wage hurt the restaurant industry?

In brief: Tipped workers are low-wage workers who need wage increases just as much as any other type of worker. Economic research does not find that boosting the minimum wage for tipped workers hurts the restaurant industry.

In detail: There is no inherent economic reason why tipped workers should be paid a lower minimum wage than other workers. The fact that U.S. law allows this can be traced directly back to racist economic practices following the abolition of slavery. Seven states do not have a separate subminimum wage for tipped workers yet still have strong restaurant and hospitality industries. Economic research on the restaurant industry finds that tipped minimum wage increases boost wages for workers without affecting employment. Similarly, when the District of Columbia increased its tipped minimum wage, the restaurant industry did not suffer in terms of employment growth or number of establishments when compared with the U.S average or nearby counties.

Don’t tipped workers earn enough to earn a living wage?

In brief: Tipped workers, including restaurant servers and bartenders, are overwhelmingly low-wage workers. Many struggle to make ends meet, especially those in states where they can be paid less than the minimum wage.

In detail: Tipped workers are more than twice as likely as non-tipped workers to be in poverty. Poverty rates of tipped workers who live in states that use the federal tipped minimum wage of $2.13 are substantially higher than poverty rates of tipped workers in states that use the same minimum wage for all workers, regardless of tips.

The subminimum wage for tipped workers exacerbates economic insecurity for many workers. Employers are legally required to ensure that on a weekly basis, tipped workers’ tips cover the gap between the tipped minimum wage and the regular minimum wage for all hours worked that week, on average. If they do not, employers are responsible for making up the difference. In practice, this requirement is exceptionally difficult to enforce, as it is largely left to workers themselves to track their hours and tips, make the relevant calculations, and then confront their employer if something seems amiss. As a result, tipped workers—who are already paid low wages—are particularly vulnerable to wage theft.

Will raising/eliminating the tipped minimum wage lead to fewer tips or force restaurants to end tipping?

In brief: Tipping is deeply embedded in U.S. culture. Even in places with no separate tipped subminimum wage, workers still receive tips and typically have higher overall take-home pay than their peers in places with a separate tipped subminimum wage.

In detail: In the seven states that do not have a tipped subminimum wage, tipped workers continue to receive tips. According to the Toast platform, California (a state where tipped workers receive the full minimum wage) had the lowest tipping rate (i.e., the average percentage tip on a bill) in the country at 17.2%. This is less than 5 percentage points less than Delaware, the highest tipping state (21.8%). By contrast, the effective minimum wage for tipped workers in California ($16.90) is more than seven times greater than in Delaware ($2.23). EPI research finds that tipped workers in states without a lower tipped subminimum wage earn, on average, 17% more per hour in total take-home pay (base wages plus tips) than tipped workers in states that use the federal $2.13 tipped subminimum.

There is nothing wrong with workers receiving tips for their work in service jobs, but formalizing tipping in minimum wage law allows employers to shift responsibility for paying their workers onto customers. This in turn means workers are more vulnerable to harassment, discrimination, and other forms of abuse. Restaurant workers, particularly women, are subject to the highest rates of sexual harassment of any industry. Research has also found that racial discrimination leads to Black workers receiving fewer tips than their white counterparts. Relying on tipping means workers have less ability to avoid or protect themselves from harmful interactions in the workplace.

Does the so-called “no tax on tips” deduction eliminate the need to raise the tipped minimum wage?

In brief: The 2025 budget tax bill did create a temporary tax deduction for tipped income, but for most tipped workers the benefits are modest and pale in comparison to the benefits of increasing the wage floor.

In detail: The 2025 Republican budget bill created a new, temporary federal income tax deduction for tipped income. This policy does little to address the precarity of tipped work and the benefits to most tipped workers pale in comparison to the gains they would receive through a significant minimum wage increase. The tax deduction encourages employers to rely more on tipped jobs and avoid raising base wages, exacerbating the low wages and challenging conditions of most tipped jobs (see Questions 9 and 10). Many tipped workers earn too little to qualify for the benefit, and those that do will likely see modest tax benefits. Whereas the average annual benefit for an eligible tipped worker will be around $1,700 a year for the three remaining years the deduction is in place, a minimum wage increase to $15 per hour would boost earnings by $3,200 a year for a full-time worker, in perpetuity.

Aren’t most minimum wage workers teenagers?

No, the vast majority of workers impacted by the minimum wage are not teenagers. Low-wage work is a widespread problem and not just isolated to younger workers. EPI’s analysis of the 2025 Raise the Wage Act found that only 14% of the workers that would be impacted by the policy were younger than 20 years old.

Will raising the minimum wage hurt young workers in their first jobs? 

In brief: Higher minimum wages cause little to no employment changes for teenagers.

In detail: The majority of studies find little to no evidence that the minimum wage causes employment losses for teen workers. Instead, their incomes increase as they earn higher pay. It is also worth keeping in mind that teen workers are a minority of low-wage workers, and a shrinking share of the workforce overall as the cultural and economic emphasis on education has grown. To that end, minimum wage increases can support young workers’ educational attainment, particularly for low-income teens. Minimum wage increases significantly improve high school graduation rates for low-income students, a vital investment that can have large long-term consequences for those workers’ lifetime earnings.

Do workers benefit from the FLSA’s subminimum wage for workers with disabilities?

In brief: The federal subminimum wage for disabled workers does not provide real wage protections and is out of step with the most effective ways to boost employment for workers with disabilities.

In detail: Under Section 14(c) of the Fair Labor Standards Act, employers can apply for special certificates with the Department of Labor that allow employers to pay workers with mental or physical disabilities less than federal minimum wage. Employers can only apply for certificates if the worker’s disability actually impairs the worker’s earning or productive capacity. An overwhelming share (96%) of 14(c) employees work in so-called “sheltered workshops” which put workers with developmental disabilities in isolated, noncompetitive environments.

These certificates apply to a small number of workers (less than 37,000 nationally) but also produce exceedingly low pay for workers with disabilities. Nearly half of 14(c) workers were paid less than $3.50 an hour, exacerbating the economic precarity experienced by disabled workers. Disabled adults are 24.1% more likely to live in poverty than other adults. The low pay permissible under 14(c) perpetuates these workers’ struggle to make ends meet. This measure is also unnecessary for providing well-paying employment opportunities to workers with disabilities.

Researchers studying state repeals of 14(c) have found that the change has not hurt disabled workers’ employment. In Maryland, eliminating the provision caused no significant change to disabled worker employment, while in New Hampshire, employment increased. An analysis of the federal AbilityOne program, which is composed of nonprofits that primarily employ workers with disabilities, found that state and local minimum wage increases did not impact the employment of those workers. Employers also do not appear to shift more workers to 14(c) certificates in response to minimum wage increases.

Overall, the use of 14(c) certificates has been declining over time. Across the country, 27 states and D.C. have eliminated or restricted the use of the provisions, reflecting that the policy does not offer real wage protections for disabled workers and that there are superior models of employment for workers with developmental disabilities. Respecting the dignity of workers with disabilities requires prioritizing real pay and inclusion in supportive, but integrated employment opportunities.

Many cities and states already have minimum wages above $15 an hour. Is increasing the federal minimum wage still important?

In brief: Yes, tens of millions of workers still earn less than $15 an hour. In many states and cities, higher wage floors are needed to provide meaningful economic security.

In detail: In the last decade, dozens of cities and states have responded to federal minimum wage inaction by enacting stronger wage floors, in many cases reaching or exceeding $15 an hour. As of 2026, more workers work in a state with at least a $15 minimum wage than in a state using the federal minimum wage. However, there are still 20 states that use the federal $7.25 wage floor and around 14 million workers earn less than $15 an hour.

Even places with recent minimum wage increases might need higher wage floors. The first $15 minimum wage was enacted in 2013. Prices have risen substantially since then, and consequently, the value of targets like $15 an hour has declined significantly. According to EPI’s Family Budget Calculator, $15 is not a living wage almost anywhere in the country. Many cities and states have at least partially protected their wage floors by adopting automatic annual adjustments to account for inflation, but if the initial value is too low, this inflation-indexing only locks in an unlivable floor.

Very few workers earn the federal minimum wage of $7.25 an hour. Is the minimum wage even relevant anymore?

In brief: The fact that so few workers earn the federal minimum wage is a policy failure, not a reason to abandon the policy. The minimum wage is a vital tool for lifting wages and addressing systematic power imbalances between workers and employers. The failure to adequately raise the minimum wage over time has left millions of workers being paid less today than they could have been earning.

In detail: It is true that a small fraction of the labor force earns exactly the federal minimum wage, but this is a policy failure that has left tens of millions of workers with lower wages. Had Congress simply raised the federal minimum wage to keep pace with inflation since the late 1960s, it would be over $12.50 today. According to EPI’s Low Wage Workforce Tracker, 14 million workers earn less than $15 an hour, while 42 million earn less than $20 an hour. The minimum wage does not just impact workers at the very bottom of the wage distribution; it exerts upward pressure for low-wage workers in general. Minimum wage increases create “spillover effects,” where workers above the new minimum wage threshold also see wage increases as employers keep wage ladders and seniority consistent in their firms.

It is important to recognize that without leveraging policy tools like the minimum wage, the low-wage labor market gives employers excess power to set low wages. Workers have limited information about the wages and work policies at alternative employers and can be constrained in their job choices by limited transportation options or the need to maintain specific schedules for child care and other family needs. These economic “frictions” add up, providing leverage for employers to pay lower wages than is optimal for the economy. In short, the longstanding failure to increase the federal minimum wage suppresses worker pay, leaving low-wage workers worse off every year there is no increase.

Does raising the minimum wage lead to automation of low-wage jobs?

In brief: The minimum wage is not a primary cause of automation of low-wage work, but automation is changing the tasks and occupations of some low-wage workers.

In detail: Increasing the cost of low-wage labor can encourage businesses to invest in automation. This can lead to disruption for specific low-wage jobs, or changes in the roles in those occupations. Since we see that the minimum wage does not increase unemployment for low-wage workers (Question 1), the effects of automation on these low-wage jobs are either limited or counterbalanced by expanding employment in other occupations. Evidence suggests that while in recent years automation is taking over a growing number of routine tasks from low-wage workers, the minimum wage has a limited contribution in driving that adoption. Researchers with access to extensive data on McDonalds franchises nationwide found that, while the adoption of touch-screen ordering kiosks grew significantly between 2017 and 2019, there was no evidence that uptake was driven by minimum wage increases. So far, the overall employment impact of automation on low-wage workers has been insignificant, as reductions in occupations with high amounts of routine tasks have been replaced with greater demand for jobs with interpersonal tasks.

Technology is a tool, but the balance of labor market power determines who it helps. Automation has been a feature of our economy since the industrial revolution, boosting productivity in our economy. Technological change can disrupt employment for specific sectors or professions, but in aggregate the economy benefits. Workers benefit from these productivity increases when there are strong labor institutions like access to unions, a strong minimum wage, and policies that support full employment. When those institutions are weak, the gains from technological advancement are not shared widely and contribute to increased inequality.

Key takeaways:

• Misclassification of workers as independent contractors is a pervasive and widespread problem. AAPI workers are overrepresented in three of the 11 commonly misclassified occupations: manicurists and pedicurists, home health aides, and personal care aides. Vietnamese, Bangladeshi, Filipino, Samoan, and other Pacific Islander workers are overrepresented within these occupations.
• Groups with lower median hourly wages also have larger shares of their working populations in the 11 commonly misclassified occupations.
• Federal protections against misclassification are limited and currently under attack by the Trump administration. The state and local landscape for curbing misclassification is varied, which leaves some workers less protected than others.

In March, EPI published updated research highlighting the cost to workers of being misclassified as an independent contractor for 11 commonly misclassified occupations. Asian American and Pacific Islander (AAPI) workers were overrepresented in three of those occupations—manicurists and pedicurists, home health aides, and personal care aides—relative to their share of the overall workforce.

Most federal, state, and local labor laws apply only to employees and not to independent contractors, so misclassification strips workers of key protections such as minimum wage laws or qualifying for employer-provided health insurance and retirement benefits. Additionally, both misclassified workers and social insurance funds lose out on income: the report conservatively estimates that for the three jobs in which AAPI workers are overrepresented, misclassification costs workers at least $7,000 annually and costs social insurance programs $600 to $800 per worker each year.

With the understanding that the umbrella term “AAPI” encompasses an immensely diverse population both in ethnic origin but also in economic outcomes, this piece goes beyond the narrow view that all AAPI workers are high-wage earners. Below, we provide more detail on which groups of AAPI workers are most likely to be employed in lower-wage commonly misclassified occupations.

Disaggregated data shed light on particular AAPI communities that may be vulnerable to misclassification

Across all occupations, AAPI workers comprise approximately 8% of the total workforce. For three of the 11 occupations highlighted in the report—manicurists and pedicurists, home health aides, and personal care aides—AAPI workers make up 67%, 13%, and 10% of employment, respectively, according to Current Population Survey (CPS) data.

Table 1 provides a detailed breakdown of the composition of the AAPI workforce for the three occupations in which AAPI workers are overrepresented. Here, we use the American Community Survey (ACS) as it offers detailed race definitions which the CPS does not offer due to sample size restrictions.

Asian Indian and Chinese populations combined make up over 40% of the working-age AAPI population, thus their relatively large shares of the AAPI workforce in these occupations are not surprising. However, several groups are disproportionately represented across these occupations compared with their share of the overall AAPI workforce.

For example, Bangladeshi workers make up 5.1% of AAPI workers employed as home health aides while only constituting 1.1% of the total AAPI workforce. Chinese workers represent almost half (47.7%) of AAPI home health aides while representing just over one-fifth of the overall AAPI workforce (20.9%). AAPI employment among manicurists and pedicurists is largely held by those of Vietnamese origin (71.4%).

Finally, a majority of AAPI personal care aides are either Filipino (32.8%) or Chinese (20.8%). Filipino workers, however, are overrepresented by twice their share of the overall workforce. While Samoans and other Pacific Islanders comprised a much smaller share of personal care aide employment, they are also overrepresented in this occupation by more than twice their share of the overall workforce.

Table 1

Figure A provides a more comprehensive picture of the share of each detailed group employed across all 11 commonly misclassified occupations, revealing that smaller communities—often overlooked because of their size relative to the aggregate AAPI workforce—may be among the most vulnerable to misclassification. Workers belonging to seven of those groups are more likely than the average U.S. worker to be employed in one of those occupations. Almost 20% of Vietnamese workers are employed in one of those occupations, with over half concentrated as manicurists and pedicurists.

Samoan, Hawaiian, and other Pacific Islanders have the next highest shares working in the 11 occupations, making up 15% or more of their total working-age population. These groups also earn lower median hourly wages than the national median and the aggregate AAPI median hourly wage. Their disproportionate representation in commonly misclassified occupations further exposes these workers to wage suppression due to misclassification.

Misclassification enforcement varies by state—meaning different AAPI populations can be disproportionately impacted

Federal protections from misclassification are limited and are currently under attack by the Trump administration, which has proposed a rule to weaken standards to determine worker classification under the Fair Labor Standards Act, the Family and Medical Leave Act, and the Migrant and Seasonal Agricultural Protection Act. The proposed rule narrows the definition of who is a covered employee under these statutes, encouraging employer schemes to reclassify their employees as independent contractors to evade those obligations.

Broadly, the Trump administration has been actively dismantling long-standing federal worker protections, leaving states to bear the responsibility of ensuring workers are given rights and protections and that they can exercise them. For most states, labor and employment protections only apply to workers classified as employees, meaning workers misclassified as independent contractors are denied their legal rights and protections.

EPI’s 2026 misclassification report outlines state and federal policy recommendations that ensure proper enforcement mechanisms to curb misclassification. One of the recommendations includes implementing the ABC test. Unlike the six-part “economic reality” test or the “common law” test, the ABC test presumes that a worker is an employee unless they can demonstrate they are an independent contractor based on three criteria. Placing the onus on the employer to determine the employment status of a worker provides protections against misclassification and extends proper protections to workers. Many states have adopted the ABC test for unemployment insurance programs and, to a lesser extent, for wage and hour orders and other employment applications.

As shown in Figure B, The AAPI population is highly concentrated across a handful of states. Almost half of the prime-age working Asian population is concentrated in California, New York, and Texas, and a majority of the Pacific Islander population resides in California, Hawaii, and Washington. Overall, 21 states have significant numbers of AAPI residents, and some are home to large shares of specific AAPI communities. For example, the Hmong community in Minnesota and the Burmese community in Indiana are concentrated in states that have smaller total AAPI populations.

The current landscape for state policy protections against misclassification is quite varied. For example, among the states with the largest AAPI populations, California is the only state to adopt the ABC test for both unemployment insurance and employment law, although certain occupations are exempt from the test—including app-based drivers. California also institutes penalties for misclassifying a worker, which can include restitution payments and, if the misclassification was willful, a penalty between $5,000 to $25,000 per violation.

Texas, on the other hand, has significantly less state enforcement. Apart from using the common law test for its unemployment insurance program and providing a definition of an independent contractor for workers’ compensation, Texas mainly relies on federal law for classifying workers as employees. In the last 15 years, Texas lawmakers have introduced several bills that would create penalties for misclassifying workers in the construction industry, but all have stalled or failed.

Figure B

Comprehensive protections are needed to protect workers from misclassification

AAPI workers are facing multi-pronged attacks from the Trump administration through the degradation of federal protections for workers, immigration, and equity. Occupational segregation and other labor market disparities lead women, people of color, and immigrants to be disproportionately represented in occupations that are commonly misclassified. These factors—in addition to historical and current geopolitical relations that shape the flow of labor to the U.S., immigration and citizenship status, and English language proficiency—can contribute to the concentration of AAPI workers in these occupations. Disaggregated data further identify which specific AAPI communities are overrepresented, revealing that smaller, less economically secure groups are often most exposed to the costs of misclassification. Strong policies at the federal, state, and local levels are needed to combat misclassification and to ensure workers can exercise their rights.

Key takeaways:

• The federal minimum wage is at its lowest real value in 77 years. Frozen at $7.25 since 2009, the federal minimum wage has lost 30% of its purchasing power during this 17-year freeze.
• Setting the federal minimum wage at two-thirds of the national median wage would raise pay for 39.6 million workers in 2030, about 1 in 4 of the wage-earning workforce.
• The policy would move the federal floor meaningfully toward one definition of a living wage, meeting EPI’s Family Budget Calculator thresholds in half of U.S. counties for a single adult working full time. But it falls short for many families, meaning that policies to strengthen unionization, provide a more robust safety net, and keep unemployment low remain essential.
• Decades of economic research support this two-thirds benchmark, finding little to no employment loss from ambitious minimum wage increases.
• Indexing the federal minimum wage to median wage growth would lock in these gains. Median wages typically outpace prices, so median wage indexing would prevent the kind of decades-long slide that has eroded the current floor.

Introduction

The federal minimum wage, frozen at $7.25 since 2009, is now at its lowest real value in 77 years and a major driver of the affordability crisis facing low-wage workers. For over a decade, the senior Democrats on the House and Senate’s labor committees have consistently introduced and championed the Raise the Wage Act, which would significantly raise the federal level (most recently, to $17 an hour in 2030) and index it to median wage growth going forward. But Congress as a whole has failed to take action on the legislation. In the absence of federal movement, states have moved on their own: Thanks in large part to the Fight for $15 campaign, 21 states and the District of Columbia, home to half of all U.S. wage earners, will have a minimum wage of at least $15 by 2028. But that patchwork still leaves 20 states—home to about 55 million workers—at $7.25, and updating the federal floor to a modern benchmark is the only way to reach these workers.

Raising the federal minimum wage to two-thirds of the national median wage would lift pay for nearly 40 million workers, about a quarter of the workforce. Two-thirds of the median—equivalent to roughly $17.70 today, a projected $20 in 2030, and a projected $25 in 2038—matches the benchmarks used in other high-income countries and tracks the direction of recent minimum wage research. Indexing to median wage growth thereafter would keep the floor from losing ground to inflation or falling behind the broader economy.

A federal minimum at two-thirds of the national median would eliminate poverty wages and move the floor meaningfully toward a living wage in much of the country: A single adult working full time could cover modest expenses in half of U.S. counties under EPI’s Family Budget Calculator thresholds. Maintaining the two-thirds minimum-to-median ratio would lock in those gains, improving affordability for U.S. workers and their families. It would also durably narrow the gap between low-wage workers and the typical worker, with Black workers and women seeing the largest benefits.

The two-thirds benchmark is also well-supported by economic research. Decades of studies of state and federal minimum wages find that higher floors raise pay for low-wage workers with little to no effect on employment, and a smaller but growing body of work on minimum wages approaching two-thirds of the median reaches the same conclusion. Setting the federal floor at two-thirds of the median, and updating it annually, would raise incomes at the bottom and prevent the kind of decades-long slide that has left the current minimum at its lowest real value in 77 years.

The outdated federal minimum wage and extent of low pay

The federal minimum wage is now at its lowest real value in 77 years. Stuck at $7.25 since 2009, it is in its longest stretch without an increase since the federal wage floor was established in 1938 (Figure A). Inflation has eroded 30% of its purchasing power over those 17 years, gradually cutting real pay for the lowest-wage workers in states still tied to the federal floor. Simply indexing the 2009 wage to inflation, a far weaker standard than this report proposes, would put the federal minimum at about $10.60 today.

The minimum wage was once meaningfully higher in real terms. Civil rights organizers in the late 1960s pressed Congress not only to raise the wage floor but also to extend its coverage to service industries that had previously been excluded because they disproportionately employed Black workers. By 1968, the federal minimum wage reached $1.60 per hour, equivalent to $12.62 in 2026 dollars and roughly 61% of the national median wage at the time, close to the two-thirds benchmark this report proposes. Even a federal minimum wage of $12.62 today would raise the wages of about 12 million workers.

Because Congress has not raised the federal floor in 17 years, states and localities have moved on their own. Thirty states, the District of Columbia, and dozens of cities and counties have raised their minimum wages, many in response to the Fight for $15 campaign (EPI 2026a). By the end of 2028, more than half of the U.S. workforce, about 74 million workers, will live in a state with a minimum wage of at least $15.

The state-by-state patchwork has delivered real successes but also left tens of millions of workers behind. Roughly 55 million people work in the 20 states still tied to the $7.25 federal floor, and they are nearly twice as likely as workers elsewhere to earn less than $15 per hour. Nationally, almost no one is paid exactly $7.25 anymore: The floor is so low it rarely binds. Yet 14 million workers, about 9% of the workforce, still earn less than $15. Closing that gap and preventing the federal floor from eroding further requires a national standard pegged to a modern benchmark.

A new standard: Two-thirds of the median wage

The federal minimum wage suffers from two related deficiencies: Its level is too low, and it does not adjust as the economy grows. Both can be solved by tying the federal minimum to two-thirds of the national median wage. Congress would first raise the floor to that level, and each subsequent year the minimum would adjust to maintain the same ratio.

First, the new benchmark replaces a poverty-level federal floor (Hickey and Cid-Martinez 2025) with one that pushes the minimum wage toward a living wage. As Oakford (2026) argues, two-thirds of the median is “a realistic stepping stone to living wages,” and standards below that ratio leave too large a gap between earnings and the cost of necessary expenses. A federal minimum at two-thirds of the median also better fulfills the original promise of the federal standard, which Congress described in 1937 as protecting “this Nation from the evils and dangers resulting from wages too low to buy the bare necessities of life” (U.S. Congress 1937).

Second, maintaining the two-thirds ratio guarantees automatic increases as the economy grows, ending the recurring erosion that comes from a frozen federal floor. Because median wages typically outpace prices, median wage indexing produces real gains, not just inflation protection. In 2025, two-thirds of the national median wage was $17.11 per hour (Figure B), and today, it is estimated to be $17.70. By 2030, applying Congressional Budget Office (2026) Employment Cost Index projections, it would reach $20.02. A federal minimum tied to two-thirds of the median would likely reach or exceed $25 by 2038, four years sooner than if a $17.11 wage in 2025 had been indexed only to the cost of living going forward.1

Nineteen states and the District of Columbia index their minimum wages to inflation, but Connecticut goes further by indexing to average wage growth and capturing the real gains that wage growth typically delivers above prices. Congress should follow Connecticut’s lead and link the federal minimum to the national median wage. Of course, indexing to price inflation would be an enormous improvement to current federal minimum wage policy and to state and local minimum wage policies that have also failed to implement automatic increases. Tying the minimum wage to median wages—i.e., indexing the minimum to typical workers’ wage growth—would yield even larger increases over time.

Tying the federal minimum to two-thirds of the median would also durably narrow inequality in the bottom half of the wage distribution. Whenever the minimum wage fails to keep pace with economy-wide wage growth, the gap between low and median earners widens. But a substantial increase in the minimum to a fixed ratio of the median shrinks and bounds that gap by construction. The gains disproportionately affect Black workers and women, who are overrepresented in low-wage jobs due to persistent racism and sexism (Banks 2019). Minimum wages are a major determinant of Black-white wage gaps (Derenoncourt and Montialoux 2020; Wursten and Reich 2023), and the long erosion of the federal minimum was a leading driver of widening pay inequality among women (Autor, Manning, and Smith 2016).

The state of minimum wage research and new policies

Minimum wages and job losses

A federal benchmark of two-thirds of the national median would significantly raise wages, and recent research strongly supports the conclusion that ambitious minimum wage targets work as intended, with little to no employment downsides. Across more than three decades of modern economic research, the median estimated employment effect is small; among studies that look at all low-wage workers rather than narrow subgroups, the effect is essentially zero (Zipperer 2024). The recurring scare stories about job losses are not borne out by the body of evidence.

Businesses adjust to higher minimum wages through what Dube (2026b) and Bernstein (2013) call the “Three P’s”: productivity, prices, and profits.2 Take productivity first. Higher wages reduce the rate at which workers quit, particularly in high-turnover sectors like restaurants and retail. That lowers hiring and training costs and means employment levels can hold steady even as new hiring slows. Better-paid workers, and workers with longer tenure, are also typically more productive, further offsetting the cost of the wage increase.

The minimum wage also redistributes income to low-wage workers when employers cover higher labor costs through reduced profits or modestly higher prices. Vergara (2026) and Coviello, Deserranno, and Persico (2022) both find that minimum wage increases shrink profits in low-wage industries. Price pass-through is small in aggregate terms because low-wage workers’ earnings are only a fraction of total labor costs, which are themselves a fraction of total business expenses. California’s $4 overnight increase in the fast-food minimum generated a one-time increase in fast-food prices of 2.1% to 3.6% (Sosinskiy and Reich 2026; Clemens et al. 2026). To put that in context: The price of a $6.00 hamburger would have risen to about $6.17.

How high is too high?

A common way to measure the level of a given minimum wage is to use the minimum-to-median wage ratio. Sometimes called the Kaitz index, the minimum-to-median wage ratio compares the minimum with the underlying distribution of wages by measuring the share of the typical wage that the floor reaches. This report proposes setting that ratio at about 67%. Most of the U.S. evidence base reflects periods when the ratio sat well below that level, because until recently, U.S. minimum wages were rarely considered high by today’s standards. But a growing body of recent research, together with recent state and local policies, has pushed the evidence into higher ratios—and the results are the same: There is substantial room for higher minimums without large employment losses.

Cengiz et al. (2019) found no negative employment effects at minimum-to-median ratios up to 59%. Dube and Lindner (2021), studying city-level minimum wages with ratios averaging 58% to 64%, found small and statistically insignificant effects. Godoy and Reich (2022) found no employment effect across localities with ratios ranging from 56% to 82%. And the 1968 federal minimum, which reached roughly 61% of the median,3 has been reexamined in two recent studies that likewise found small or no employment effects (Bailey, DiNardo, and Stuart 2021; Derenoncourt and Montialoux 2021).

The most direct evidence that the floor can go meaningfully higher comes from California’s $20 fast-food minimum wage. In April 2024, the state raised the wage for fast-food chain workers from $16 to $20, pushing the ratio of that minimum to the state’s median wage to about 74%, well above most U.S. precedents.4 One might worry that customers would substitute toward lower-priced independent restaurants exempt from the policy, generating job losses at the chains. The actual evidence shows otherwise. Despite the large wage increase, research finds little to no employment effect of the policy (Bivens and Zipperer 2026), and the median employment effect in Dube (2026a) is essentially zero. Evaluations of the UK minimum wage through 2019, when it reached nearly 60% of the median wage, also find small, statistically insignificant effects on the employment of low-wage workers (Giupponi et al. 2024).

A federal benchmark set at two-thirds of the national median will push some states above two-thirds of their own median wage. There is good reason to be optimistic about employment changes there as well. The studies above already span a wide range of Kaitz ratios, from the high 50s through the low 80s, and consistently find little or no employment effect. California’s $20 fast-food minimum extends this evidence to a 74% state-level ratio with essentially no employment losses, and only five states would have a minimum-to-median wage ratio above that threshold under the proposed federal benchmark. And while the minimum wage would be at a higher level relative to the state median in those states, it would still be less than a “living wage” for many families in those areas, as I discuss later.

Even if some employment loss does occur, that is not the right test of policy success. Low-wage labor markets are dominated by job-to-job churn, so reduced employment in response to a higher minimum typically shows up as longer gaps between jobs rather than workers permanently shut out of the labor market (Cooper, Mishel, and Zipperer 2018). On net, low-wage workers come out ahead in annual earnings when significantly higher hourly pay more than offsets a modest increase in unemployment.

Policymakers and organizers campaigning for minimum wage increases have considerable room to maneuver above the current federal floor before needing to worry about job losses. To assuage concerns about employment impacts, a federal proposal could be structured to limit annual increases of the federal minimum wage so that they never exceed two-thirds of the national median wage. States and localities, of course, can and should continue to push for higher minimum wages, as many will have higher median wages and costs of living than the national average.

Existing proposals and policies reaching two-thirds of the median

Some recent federal proposals already target or are consistent with the two-thirds benchmark. The recent Give America a Raise Act would raise the federal minimum to $20 by 2029, close to this report’s projection of two-thirds of the 2029 median wage ($19.44). The Living Wage for All Act names the two-thirds benchmark explicitly and locks in indexation in statute: “once the minimum wage equals two-thirds of the national median hourly wage, it shall thereafter be automatically adjusted each year to maintain that ratio.” The Bold Economic Program for America (Reich 2026) likewise proposes $20 by 2030, and Oakford (2026) embeds the two-thirds target in a broader portfolio that includes just cause protections and stronger wage theft enforcement.

The benchmark also aligns U.S. policy with international practice. The UK Low Pay Commission has targeted two-thirds of the median for the National Living Wage since 2024, and in the EU, 17 of 22 countries benchmark their statutory minimum wages to a ratio of the median or average wage. The 2022 European Union Minimum Wage Directive obligates member states to use “indicative reference values”—such as 60% of the gross median wage—to assess adequacy of their wage standards (Luebker and Schulten 2026).

Some of these international benchmarks may look numerically lower than two-thirds, but they are usually defined against a different denominator. Germany, for instance, benchmarks against the median wage of full-time workers. In the United States, the full-time median is about 10% higher than the overall median, so 60% of the full-time median is roughly equivalent to two-thirds of the overall median that this report proposes.

Implementing a minimum wage equal to two-thirds of the median wage

Any federal legislation will need a phase-in period, but it must specify two things: a clear path to the target and an explicit guarantee that automatic median wage indexing kicks in once the floor reaches two-thirds of the median.

Implementing the benchmark also requires choosing a wage source. Legislation should designate the Department of Labor (DOL)—which already publishes median wage estimates through the Bureau of Labor Statistics (BLS) Occupational Employment and Wage Statistics (OEWS)—to publish the official median wage each year. DOL has two ready sources: OEWS, an establishment survey, and the Current Population Survey (CPS), the household survey already used to produce the unemployment rate, which collects detailed wage and hours data.

Each source has tradeoffs. The CPS is timelier, with wage data available at a one- to two-month lag, but smaller samples, the difficulty of computing hourly earnings for salaried workers, and respondents’ tendency to round wages all introduce noise. OEWS uses an established BLS hourly wage methodology and median wage calculation that may be less volatile, but it is published with a one-year lag and pools data from earlier, lower-wage years. DOL could pick one source or use a weighted average; recent data show only about a $1 difference between the two surveys.5 Either way, expanding resources at BLS and the Census Bureau would strengthen the underlying data and support further refinements to the methodology.

Once DOL has a baseline median, indexing requires projecting that median forward to the year the new minimum takes effect. The UK Low Pay Commission, which recommends a two-thirds median target to the UK government, offers a useful template: It estimates a midyear median wage by combining lagged historical data with timely indicators and short-run forecasts (Low Pay Commission 2024). A concrete schedule illustrates the approach. To set the minimum for January 1, 2030, DOL would announce the new wage on July 1, 2029, six months in advance, based on its best projection of the July 2030 median, the midyear point representative of the median wage workers will face on average throughout 2030. The projection would proceed in three steps: compute the 2028 median from CPS or OEWS data; roll it forward to early-to-mid-2029 using a combination of available data—like the Current Employment Statistics, the Consumer Price Index, and the Employment Cost Index (ECI); and then roll it forward one more year using short-run wage projections like those in the CBO Budget and Economic Outlook (2026).

National and state effects of a federal minimum wage at two-thirds of the median

To estimate the economic benefits of a federal minimum wage set at two-thirds of the median wage, I model how many low-wage workers would see higher pay under this policy. I assume the policy is phased in over five years, so that if it went into effect today, the federal minimum would reach two-thirds of the median in 2030 and then automatically adjust each year to maintain that ratio.

Concretely, I assume the federal minimum rises to $12 immediately in 2026 and then increases incrementally to $20 in 2030, which is about two-thirds of the projected national median wage.6 Legislation should build in a path adjustment if 2030 median wages come in higher or lower than projected.

I focus on the effects in 2030. I assume the same phase-in path and automatic indexing applies to the federal tipped minimum wage, which has been frozen at $2.13 per hour since 1991.7 Wages elsewhere are assumed to grow in line with CBO ECI projections from 2026 to 2030, and the model incorporates the effects of scheduled state-level minimum wage increases (see the appendix for details).

In 2030, a federal minimum wage equal to two-thirds of the national median would raise pay for 39.6 million workers, about 1 in 4 of the wage-earning workforce (Table 1).8 Annual earnings would rise substantially, and the gains would be largest for Black workers: A full-time, full-year Black worker affected by the increase would earn about $5,000 more per year, compared with $4,400 for all affected workers. In line with other minimum wage increases, women would gain more than men, with 31% of women seeing higher pay compared with 23% of men.

Adults ages 20 and over would make up 9 in 10 affected workers (teens would have the largest share of affected workers of any age group, but they make up a small share of total employment). The problem of low pay is far from limited to the youngest workers.

Although several states with scheduled minimum wage increases will see the gap between their minimum wage and the new federal floor shrink, workers in every state would still be affected (Figure C). With the exception of D.C., no state’s scheduled 2030 minimum wage reaches $20. The largest gains go to workers in the 20 states still tied to $7.25: 1 in 3 (33%) would see a raise, and annual pay for those affected would rise by about $6,200 on a full-time, full-year basis.

Figure C

Beyond who gets a raise and how much, a related question is whether the raise is enough to cover a family’s basic costs. Setting a target of two-thirds of the median would push the federal floor much closer to a living wage for many families. What counts as a living wage varies across the country, depending on local costs and on family size and composition. EPI’s Family Budget Calculator thresholds make this concrete: They calculate the income a given family type needs in a given place to afford a “modest but adequate” standard of living. For example, in 2025 a two-adult, one-child family in Los Angeles County, California, needed about $118,000 in annual pretax earnings to pay for housing, food, child care, transportation, health care, taxes, and other necessities. In more rural Early County, Georgia, a similar family needed about $74,000. These budgets are minimal by design, with no allowance for savings, emergencies, retirement, college, or entertainment.

Of course, other business income and government provided social benefits can lower the amount of labor market earnings a family needs to maintain the same standard of living. Gould, Mokhiber, and deCourcy (2024) report that, according to Congressional Budget Office data, about 81% of a middle-income family’s budget is met by labor market income.

Applying that 81% adjustment to EPI’s Family Budget Calculator thresholds gives a more useful benchmark for what wages need to deliver. Under this adjusted measure, the Los Angeles County family would have needed about $96,000 in earnings to meet their 2025 budget, equivalent to both adults working full time, year round at about $23 per hour. The Early County family would have needed about $60,000, equivalent to both adults working full time, year round at $15 per hour.

The implications for everyday affordability are significant. After inflating these adjusted thresholds to 2030 dollars using CBO Consumer Price Index projections, a federal minimum at two-thirds of the median substantially closes the gap between full-time annual earnings and necessary expenses for low-wage workers and their families. A single adult working full time at the 2030 minimum of $20 would cover modest but adequate expenses in half of U.S. counties. Two full-time working parents with two children would meet their family budget in roughly a quarter of U.S. counties. Using similar thresholds, Oakford (2026) finds that a federal minimum at two-thirds of the national median in 2025 would be “90% to 99% of the median living wage of a single adult without children in 16 states.”

A federal minimum tied to two-thirds of the national median would therefore make enormous progress in increasing affordability and helping families make ends meet. Closing the remaining gaps will require a broader set of policies, including strengthening other labor standards like overtime protections and their enforcement, a more robust safety net, expanded public goods like universal health insurance, fewer barriers to unionization, and a renewed commitment to full employment.

Conclusion

The federal minimum wage is at its lowest real value in 77 years, and tens of millions of low-wage workers are paying for that erosion every paycheck. Pegging the federal floor to two-thirds of the national median wage, and maintaining that ratio, would correct the two flaws that have left the floor unfit for purpose: a level too low to function as a meaningful wage standard and a structure that does not adjust as the economy grows.

A federal minimum at two-thirds of the median would raise pay for nearly 40 million workers in 2030, deliver the largest gains to Black workers and to women, and bring the floor close to a living wage in much of the country. Decades of research, recent state and local experience, and California’s $20 fast-food minimum all point to the same conclusion: The labor market can absorb minimum wages of this size with little to no employment cost. The benchmark also brings U.S. policy into line with the UK and most of the EU, where two-thirds-style targets are now standard practice.

A higher floor cannot, on its own, guarantee economic security for working people. That will require a broader agenda: a stronger safety net, expanded public goods, fewer barriers to unionization, and a renewed commitment to full employment. But updating the federal minimum to a modern, indexed benchmark is the single most direct step Congress can take to raise wages at the bottom, and the only step that reaches the 55 million workers in the 20 states still stuck at $7.25.

Appendix

State benefits

Methodology

Underlying wages are based on the 2025 Current Population Survey and between 2025 and 2030. I assume wages increase at the rate of CBO (2026a) ECI projections and because of future state-level minimum wages. Due to the Trump administration’s immigration policies and the possibility of continued labor market weakening, how the baseline level of employment grows over the next five years is very uncertain. For simplicity, I hold the estimated employment level constant between 2025 and 2030 instead of making additional assumptions about either employment rates or population growth. In terms of total population growth, this assumption may not be too far off the mark of current projections; CBO (2026b), for example, estimates that between 2025 and 2030, the civilian population ages 16 to 64 will only grow by 0.06%, or 130,000 people.

Affected workers include those “directly” affected, whose wages would otherwise be less than the new federal minimum wage, as well as “indirectly” affected workers who earn up to 115% of the new minimum (Cooper, Mokhiber, and Zipperer 2019). These particular estimates may overstate the number of workers affected because while they incorporate already scheduled state-level increases, they exclude city-level minimum wage increases, and some cities will have higher than $20 minimum wage standards in 2030. On the other hand, if the labor market continues to weaken, low-wage workers will, in the absence of minimum wages, face slower than usual wage growth because their wage growth slows disproportionately when unemployment is higher (Bivens and Zipperer 2018).

Notes

1. These projections follow CBO ECI and CPI projections from 2025–2036, and for subsequent years assume 2026–2035 annual growth rates of 2.26% for CPI and 2.94% for ECI. 

2. For additional discussion, see Dube and Lindner (2025), Schmitt (2013), and Zipperer (2023).

3. OECD (2026) estimated the U.S. minimum-to-median wage for full-time workers was 55.05% in 1968. Assuming a 10% premium for the full-time median wage relative to the overall median wage results in a minimum-to-median wage ratio of 60.56%.

4. In April 2024, the state raised the wage for fast-food chain workers from $16 to $20, pushing the ratio of that minimum to the state’s median wage to about 74%, well above most U.S. precedents.

5. The 2024 OEWS national median wage (which is the latest available data) was $23.80 (BLS 2025). The EPI State of Working America Data Library (EPI 2026b), which uses CPS wage data and which we use for wage levels throughout this paper, reports a 2024 median wage of $24.87.

6. The exact schedule simulated below is $12 in 2026, $13.50 in 2027, $15.50 in 2028, $17.50 in 2029, and $20 in 2030.

7. Ideally other subminimum wages would be phased out, including those for some workers with disabilities and youth workers. I do not model the effects of those changes due to data constraints.

8. I concentrate on the effects in 2030. Were low-wage workers’ wages to grow slower (or faster) than median wages, these estimates would understate (or overstate) the effects in later years.

References

Banks, Nina. 2019. “Black Women’s Labor Market History Reveals Deep-Seated Race and Gender Discrimination.” Working Economics Blog (Economic Policy Institute), February 19, 2019.

Bernstein, Jared. 2013. “If Increasing the Minimum Wage Doesn’t Cost Jobs, How Does It Get Absorbed?” Huffington Post, August 14, 2013.

Bivens, Josh, and Ben Zipperer. 2018. The Importance of Locking in Full Employment for the Long Haul. Economic Policy Institute, August 2018.

Bureau of Labor Statistics (BLS). 2025. “Occupational Employment and Wage Statistics (OEWS) Tables, May 2024” [Excel file], Occupational Employment and Wage Statistics. Accessed May 13, 2026. 

Clemens, Jeffrey, Olivia Eduwards, Jonathan Meer, and Joshua D. Nguyen. 2026. “The Effects of Calfornia’s $20 Fast Food Minimum Wage on Prices.” National Bureau of Economic Research Working Paper no. 34990, March 2026.

Congressional Budget Office (CBO). 2026a. The Budget and Economic Outlook: 2026 to 2036. February 11, 2026.

Congressional Budget Office (CBO). 2026b. The Demographic Outlook: 2026 to 2056. January 7, 2026.

Cooper, David, Zane Mokhiber, and Ben Zipperer. 2019. Minimum Wage Simulation Model Technical Methodology. Economic Policy Institute, February 2019. 

Coviello, Decio, Erika Deserranno, and Nicola Persico. 2022. “Minimum Wage and Individual Worker Productivity: Evidence from a Large US Retailer.” Journal of Political Economy 130, no. 9: 2315–2360.

Derenoncourt, Ellora, and Claire Montialoux. 2021. “Minimum Wages and Racial Inequality.” Quarterly Journal of Economics 136, no. 1 (February): 169–228.

Dube, Arindrajit. 2026a. “Labor Market Effects of California’s $20 Fast-Food Minimum Wage.” National Bureau of Economic Research Working Paper no. 35171, May 2026.

Dube, Arindrajit. 2026b. The Wage Standard: What’s Wrong in the Labor Market and How to Fix It. New York: Penguin Random House.

Dube, Arindrajit, and Attila S. Lindner. 2024. “Minimum Wages in the 21st Century.” In Handbook of Labor Economics, Volume 5, edited by Christian Dustmann and Thomas Lemieux. 

Economic Policy Institute (EPI). 2026a. Minimum Wage Tracker. Accessed May 5, 2026.

Economic Policy Institute (EPI). 2026b. State of Working America Data Library. Accessed May 5, 2026. 

Giupponi, Giulia, Robert Joyce, Attila Lindner, Tom Waters, Thomas Wernham, and Xiaowei Xu. 2024. “The Employment and Distributional Impacts of Nationwide Minimum Wage Changes.” Journal of Labor Economics 42, no. S1: S293–S333.

Hickey, Sebastian, and Ismael Cid-Martinez. 2025. “The Federal Minimum Wage is Officially a Poverty Wage in 2025.” Working Economics Blog (Economic Policy Institute), April 28, 2025.

Low Pay Commission. 2024. “What Will the Minimum Wage Be Next Year?” (blog post). April 19, 2024.

Oakford, Patrick. 2026. Federal Employment Standards as the Foundation of Economic Security: Revisiting Minimum Wage, Just Cause, and Tools to Combat Wage Theft. Roosevelt Institute, May 2026.

Organisation for Economic Co-operation and Development (OECD). 2026. OECD Data Explorer. Accessed May 5, 2026.

Reich, Michael. 2026. The Unexpected Effects of a $20 Federal Minimum Wage. Center on Wage and Employment Dynamics, February 2026.

Reich, Michael, and Denis Sosinskiy. 2026. “Effects of a $20 Minimum Wage: Evidence from Granular Data on Wages, Employment, and Prices.” Institute for Research on Labor and Employment Working Paper, April 1, 2026.

U.S. Congress. Senate. Committee on Education and Labor. 1937. Fair Labor Standards Act of 1937. S. Rep. No. 884, 75th Cong., 1st Sess. Washington, DC: Government Printing Office.

Wursten, Jesse, and Michael Reich. 2023. “Racial Inequality in Frictional Labor Markets: Evidence from Minimum Wages.” Labour Economics 82, 102344.

Introduction

In 2025, unionization in the United States grew to its highest levels since 2009 (McNicholas, Poydock, and Shierholz 2026). This growth is a testament to the fact that Americans increasingly view unions favorably and recognize them as critical instruments for building a just economy. Yet more than 50 million nonunion workers would join a union but are unable to do so because our nation’s labor laws allow employers to derail workers’ unionization efforts (McNicholas et al. 2019).

It is well documented that employers often hire union avoidance consultants to dissuade and weaken workers’ unionization efforts. These consultants work to prevent a union election from taking place—and if that fails, to ensure that workers vote against the union and then stall negotiations over a first collective bargaining agreement. Over the past several decades, large law firms have developed substantial business specializing in union avoidance services. These firms now play a significant role in denying workers their rights to a union and collective bargaining (Kaufman and Stephan 1995).

The role of these law firms in defeating workers’ organizing campaigns and frustrating workers’ attempts to reach a first contract has largely gone unexamined. While employers are required to disclose money spent on lawyers engaged in persuading employees on their union and collective bargaining rights, there is an exemption around reporting money spent on “advice” services, which is ill-defined under the law. Union avoidance law firms have taken full advantage of this reporting loophole and have constructed an industry providing counsel on union busting. Further, many union avoidance law firms provide employers services beyond these persuader activities, including representation at the NLRB and the stalling of first contract negotiations. 

In this report, we examine the union avoidance industry and the law firms that play integral roles in this business. We calculate the revenue law firms generate from employers who try to avoid unions and undermine collective bargaining with their workers. Further, we discuss the impacts of the union avoidance industry on workers’ ability to organize and what it means for workers, our economy, and our democracy.

Employers spend millions on union avoidance consultants

When workers seek to form a union, employers often hire union avoidance consultants to dissuade and weaken workers’ unionization efforts. These consultants include both non-attorney consultants and attorney consultants. Under the Labor–Management Reporting and Disclosure Act (LMRDA), employers and the consultants they hire must file disclosure reports on agreements in which the consultant is engaging in union-busting activities. Table 1 lists just a few of the employers who filed mandatory reports with the Department of Labor during 2025.

Table 1

These reports represent only a fraction of the total money spent on anti-union campaign services, not to mention legal counsel, representation, and litigation aimed at union avoidance. That’s for two main reasons: 1) consultants are not required to report activity that counts as “advice,” which is ill-defined but currently interpreted to exempt nearly all activities that don’t involve direct contact with workers, even though this accounts for the vast majority of work that consultants engage in; and 2) even activities that clearly must be reported very often are not. Research from LaborLab found that 57% of employers who were known to owe a financial disclosure for having hired a union avoidance consultant in 2024 had failed to file their required disclosure by June 30, 2025, three months after the filing deadline (LaborLab 2025). In 2024, a total of 153 employers filed a financial disclosure, according to the LaborLab report. This showcases a significant amount of underreporting from employers when one considers that over 3,200 union election petitions were filed in 2024, and that 71%–87% of employers hire a union avoidance consultant when faced with a union-organizing drive (NLRB 2026; DOL n.d.). If most “advice” provided by consultants were included, EPI estimates employers spend $442 million per year on both attorney and non-attorney consultants for anti-union campaign services.1

However, that still represents only a fraction of what employers spend on union avoidance. The EPI estimate excludes spending on legal counsel, representation and litigation aimed at defeating organizing drives and stalling contract negotiations, as well as strike preparation and strike-breaking services (McNicholas et al. 2019). It further excludes spending on consultants to implement or enhance employee engagement and “positive employee relations” programs that center around “union-substitution” policies (Levine et al. 2025). These programs feature techniques that are deliberately crafted to preempt, detect, and rapidly quash union organizing, including supervisor training, manipulative communication policies, surveillance techniques, “voice” mechanisms (like suggestion boxes), and employee-involvement programs (such as employee committees and teams).

As mentioned, EPI estimates that employers spend at least an estimated $442 million on anti-union campaign services provided by consultants that are designed to persuade or intimidate workers into voting “no” in union elections. Many of these consultants are also practicing attorneys who simultaneously will provide legal counsel and representation services related to NLRB proceedings. These attorneys also will help employers bend the law to their advantage during contract negotiations, prepare for and break strikes, file unfair labor practice charges to weaken unions and defend employers against such charges, sometimes appealing them not just to the NLRB but also into federal courts. Inclusive of all of these services, the traditional labor relation practices of these law firms generate an estimated $1.48 billion on average.2 When we account for overlap (much labor practice revenue comes from providing anti-union campaign services, not just representation and counsel), these two figures suggest that total spending on attorneys (whether for representation, consulting, or both) and non-attorney consultants is roughly $1.7 billion a year.3 Table 2 shows top law firms’ share of cases at NLRB and the estimated revenue the labor relations practices of these firms generated in 2024.

Table 2

Union avoidance law firms

Prominent law firms—such as Littler Mendelson, Morgan Lewis, and Jackson Lewis—have generated substantial business in union avoidance work on behalf of U.S. employers seeking to frustrate worker organizing and collective bargaining. As shown in Table 2, these law firms do a great deal of business before the National Labor Relations Board, the independent agency charged with enforcing the National Labor Relations Act (NLRA). The NLRA is the nation’s fundamental labor law that guarantees most private-sector workers the right to organize and the right to collective bargaining. However, decades of federal policy and court decisions have weakened the NLRA (Shierholz et al. 2024). Union avoidance consultants and law firms have long exploited the law’s significant loopholes, making it harder and harder for workers to win unions. For nearly 80 years, policymakers have failed to address the NLRA’s weaknesses and restore meaningful union and collective bargaining rights to workers.

These law firms have represented employers in fighting against some of the largest organizing efforts over the last decade, including worker organizing drives at Amazon, Starbucks, and Trader Joe’s (Logan 2025). These law firms have essentially created a specialized practice of union busting and together have generated billions of dollars in revenue, as shown in Table 2. The firms range from exclusively labor and employment firms to full-service corporate firms offering representation in a range of matters. The following are profiles of three law firms that have been at the center of the largest union avoidance campaigns in recent years.

Littler Mendelson

One of the largest union avoidance law firms is Littler Mendelson, a global management-side law firm with more than 1,800 attorneys who can make upwards of $1,700 an hour (Littler Mendelson 2026).4 In Littler Mendelson’s 80-year history, it has represented the likes of Amazon, Delta Airlines, and McDonald’s and has played a predominant role in Starbucks’s anti-union campaign (Logan 2022; Logan 2025). Beyond offering their union-busting services to employers, Littler Mendelson has expanded their services to include promoting anti-worker legislation. For example, Littler Mendelson’s Workplace Policy Institute (WPI) played a predominant role in opposing California’s Assembly Bill (AB) 5, legislation aimed at protecting workers by combatting misclassification (Poydock 2020). WPI also supported the passage of Proposition 22, which exempted gig workers from AB5 (McNicholas and Poydock 2019). WPI is part of the Coalition for Workplace Innovation, which has lobbied for proposals that weaken workers’ rights, including the exclusion of gig/app-based workers from employee status (Pinto 2022).

Morgan Lewis

Morgan Lewis also has a large practice aimed at union avoidance (Morgan Lewis 2026). The firm is a global law firm with nearly 2,000 attorneys, representing the likes of Amazon, REI, and McDonald’s. In addition to being one of the largest union avoidance law firms, Morgan Lewis is also known as one of the most expensive firms, with partners making $1,100 to $1,900 an hour.5 Morgan Lewis is the lead law firm engaged in the legal challenge to have the NLRB declared unconstitutional, despite employing multiple former NLRB officials (Rhinehart and McNicholas 2024).

Jackson Lewis

Another law firm with a significant union avoidance practice is Jackson Lewis, a national labor and employment law firm with a nearly 70-year history in union avoidance (Jackson Lewis 2026). The firm has over 1,000 attorneys who can make upwards of $730 per hour.6 Jackson Lewis has an especially robust presence in the higher education and health care industries but also serves major companies in a wide range of other industries, such as ExxonMobil, Amazon, and Google. As with other union avoidance law firms, Jackson Lewis’s services go beyond legal representation—often providing employers a “full-service” campaign in which they train supervisors and design materials, including speeches, to dissuade workers from organizing a union (Correia 2019).

How union avoidance law firms frustrate worker organizing

The NLRB election process is designed to be straightforward. Workers seeking to form a union file an election petition with the NLRB with signatures of at least 30% of the proposed bargaining unit. If parties cannot agree on a bargaining unit and election logistics, the NLRB will hold a hearing on issues of disagreement and then issue a decision and direct that an election be held. Either party can file post-election objections over the conduct of the election and other issues. Once these issues are resolved, if a majority of workers casting valid ballots in the election vote for union representation, the NLRB will certify the union and direct the parties to begin bargaining. 

While the NLRB election process is supposed to be relatively simple, the strategy of union avoidance law firms follows a standard playbook—they use their overwhelming resources to exploit the NLRB’s administrative processes and sometimes create nearly endless delays. This includes challenging bargaining units and election results and filing endless appeals of adverse decisions (See Appendix Table 1 for examples). The result is to create an unnecessarily complicated and protracted legal process for workers. The NLRB’s own performance objectives aim to ensure that the median age of representation and unfair labor practice cases before the Board is 180 days or less (NLRB 2025). While the NLRB has achieved this goal for many years, the median age for cases is over 100 days and for some workers, it can take years. For example, the NLRB only recently ordered Amazon—an employer known for hiring Littler Mendelson, Morgan Lewis, and Ogletree Deakins—to bargain with workers who voted to unionize over four years ago (Bensinger 2026).

Impact of union avoidance

The roughly $1.7 billion U.S. employers spend each year on anti-union law firms and consultants has an undeniable impact on workers’ ability to organize and bargain collectively. It also contributes to the creation of an economy marked by inequality: It has been well documented that the decline in unionization has contributed to increased income inequality over the last several decades (Bivens et al. 2023). It is no coincidence that the overall decline in unionization follows decades of federal policy neglect that have weakened U.S. labor law. The loopholes in U.S. labor law, which union avoidance consultants and law firms exploit, routinely frustrate workers’ organizing and collective bargaining, enabling wealthy corporations to prosper at workers’ expense.

Why would these corporations want to frustrate workers’ organizing? Consider the benefits unions provide for workers and their communities. When workers join together in a union and engage in collective bargaining, they see higher wages and better benefits (McNicholas, Poydock, and Shierholz 2026). Further, in communities with higher union density rates, working families have higher incomes, greater access to health care, and few voter restrictions (McNicholas et al. 2025). It is clear that when unions are strong, workers have more power and their communities thrive.

Despite the erosion of U.S. labor law and the standard playbook of union avoidance, workers do win unions and union contracts. In 2025, 16.5 million workers in the United States were represented by a union—an increase of 463,000 from 2024 and the highest number of unionized workers in the U.S. in 16 years. The 2025 rise in union density coincides with a high public favorability toward unions, with nearly 70% of people in the U.S. viewing unions favorably (Brenan 2025). Further, research from the Pew Research Center finds that most people in the U.S. see the decline in union density as bad for the country (60%) and bad for working people (62%) (Van Green 2025).

To sustain the modest gains seen in union density in 2025, policymakers must act to restore workers’ rights to a union and collective bargaining. This is critical to the health of our economy and to ensuring that workers receive a fair share of the profits they help produce. Policymakers must pass the Richard L. Trumka Protecting the Right to Organize (PRO) Act, which would help restore private-sector workers’ ability to form unions and bargain collectively.7 The PRO Act addresses many of the major shortcomings with U.S. labor law by establishing civil penalties for employers who violate workers’ rights, creating an election process that limits employer interference, and establishing a bargaining process for reaching a first contract in a timely manner. The PRO Act also would shed light on the union avoidance industry by requiring prompt disclosure of union-busting activities and closing the “advice” loophole through which employers and consultants have evaded reporting (McNicholas, Poydock, and Rhinehart 2021).

Acknowledgments

The authors would like to thank Joe Fast and Hannah Faris for their research assistance for this report.

Appendix

Methodology for labor practice revenue estimate

Estimated revenue of company-level labor practices and of U.S. labor practices as a whole was calculated in the following manner. 

First, we divided the number of attorneys listed in a company’s labor practice in 2026 by the number of attorneys that Law.com reported that the firm had in 2024, the most recent year for which Law.com data are available. We treated that figure as an initial indicator of the fraction of the firm’s total revenue that came from its labor practice. We then multiplied that fraction by the company’s total 2024 revenue, as reported by Law.com. Next, we discounted the result by 50%, on the conservative assumption that half of the revenue generated by attorneys in a company’s labor practice was earned for work performed in other areas of law than labor law. (Many labor relations attorneys belong to multiple practices, often practicing both labor law and employment law at the same company.) This calculation yielded our estimate of the revenue generated by a firm’s labor practice in 2024, inclusive of both representation and consulting services. 

We performed this calculation for the six law firms with 1.5% market share or more in 2024, where market share is defined here as a firm’s share of all NLRB cases in 2024. We then estimated the total revenue generated by all U.S. labor practices by dividing the sum of the six firms’ estimated labor practice revenue by the sum of the six firms’ market share.

Market share data were obtained through a custom query of NLRB data compiled by Labor Data (https://labordata.bunkum.us/). The number of attorneys in a company’s labor practice was obtained by tallying the number of attorneys listed on each company’s labor relations practice page in March 2026 and weeding out any attorneys practicing outside the U.S. 

Note: The share of revenue generated by attorneys in a labor practice that comes exclusively from labor relations services (rather than other areas of practice, such as employment law) may vary significantly by each law firm. For example, our labor practice revenue estimate for Littler Mendelson is lower than our estimate for Ogletree, Deakins, Nash, Smoak & Stewart, even though the former has greater market share than the latter does. This may be because our 50% assumption is too low in Littler Mendelson’s case. Perhaps attorneys in Littler Mendelson’s labor practice specialize in labor relations more often and more intensively than attorneys in Ogletree’s labor practice, thereby leading to higher labor practice revenue for Littler than our estimate suggests.

Appendix Table 1

Notes

1. See Celine McNicholas, Margaret Poydock, Julia Wolfe, Ben Zipperer, Gordon Lafer, and Lola Loustaunau, Unlawful: U.S. Employers Are Charged with Violating Federal Law in 41.5% of All Union Election Campaigns, Economic Policy Institute, December 2019. To arrive at the $442 million figure, we take the $338 million dollar estimate from McNicholas et al. 2019, which covered the four-year period 2014–2017, and adjust it for inflation to 2025 dollars, according to Consumer Price Index (CPI-U) estimates using the annual average of the BLS CPI-U for 2014–2017 and BLS C-CPI-U for 2025. The estimated rates for consultants are from McNicholas et al. 2019.

2. Full methodology for this calculation can be found in the methodology section in the appendix.

3. We assume about half ($221 million) of the $442 million goes to attorney consultants for anti-union campaign services, which we also capture in the law firms’ labor practice revenue of $1.48 billion. To get to the $1.7 billion, we add the remaining of the $442 million ($221 million) on non-attorney consultants with the law firm revenue estimates ($1.48 billion).

4. Author’s analysis of public court documents and engagement letters sourced from LexisNexis and municipality websites.

5. Author’s analysis of public court documents and engagement letters sourced from LexisNexis and municipality websites.

6. Author’s analysis of public court documents and engagement letters sourced from LexisNexis and municipality websites.

7. Richard L. Trumka Protecting the Right to Organize Act of 2025, [H.R. 20], 119th Cong. (2025).

References

Bensinger, Greg. 2026. “Amazon Must Negotiate with Staten Island Warehouse Workers, NLRB Says.” Reuters, April 2, 2026.

Bivens, Josh, Celine McNicholas, Margaret Poydock, Jennifer Sherer, and Monica Leon. 2023. What to Know About This Summer’s Strike Activity. Economic Policy Institute, August 2023.

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

Correia, David. 2019. “Union Busting on Campus: Jackson Lewis and Higher Education Anti-Unionism.” Verso Books, March 11, 2019.

Department of Labor (DOL). n.d. Persuader Agreements: Ensuring Transparency in Reporting for Employer and Labor Relations (fact sheet). n.d.

Department of Labor, Office of Labor–Management Standards (OLMS). 2026. “OPDR–LM-10 Employer” (web page). Accessed May 15, 2026.

Gregg, Forest. 2026. “Labor Data.” Accessed May 15, 2026.

Jackson Lewis. 2026. “About Us” (web page). Accessed May 8, 2026.

Kaufman, Bruce E., and Paula E. Stephan. 1995. “The Role of Management Attorneys in Union Organizing Campaigns.” Journal of Labor Research 16 (December 1995): 439–454. https://doi.org/10.1007/BF02685719.

LaborLab. 2025. One-Sided Transparency: The Growing Gap Between Required Annual Union Versus Employer and Persuader Filings and OLMS Compliance Efforts Continues to Widen. July 2025.

Law.com. 2026. “Fisher Phillips” (web page). Accessed May 15, 2026.

Law.com. 2026. “Jackson Lewis” (web page). Accessed May 15, 2026.

Law.com. 2026. “Littler” (web page). Accessed May 15, 2026.

Law.com. 2026. “Morgan Lewis” (web page). Accessed May 15, 2026.

Law.com. 2026. “Ogletree Deakins (web page). Accessed May 15, 2026.

Law.com. 2026. “Seyfarth” (web page). Accessed May 15, 2026.

Levine, Jonathan O., Tanja L. Thompson, Brooke E. Niedecken, and Brendan Fitzgerald. 2025. Littler’s 2025 Labor Survey Report. Littler Mendelson, September 30, 2025.

Littler Mendelson. 2026. “Our Firm History” (web page). Accessed May 8, 2026.

Logan, John. 2022. “Not Your Father’s Anti-Union Movement: Ten Key Facts About Starbucks’ Union Avoidance Law Firm, Littler Mendelson.” The Labor and Working-Class History Association (LAWCHA), March 7, 2022.

Logan, John. 2025. Corporate Union Busting in Plain Sight: How Amazon, Starbucks, and Trader Joe’s Crushed Dynamic Grassroots Worker Organizing Campaigns. Economic Policy Institute, January 2025.

McNicholas, Celine, and Margaret Poydock. 2019. How California’s AB5 Protects Workers from Misclassification (fact sheet). Economic Policy Institute, November 14, 2019.

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. 

McNicholas, Celine, Margaret Poydock, Julia Wolfe, Ben Zipperer, Gordon Lafer, and Lola Loustaunau. 2019. Unlawful: U.S. Employers Are Charged with Violating Federal Law in 41.5% of All Union Election Campaigns. Economic Policy Institute, December 2019.

Morgan Lewis. 2026. “Our Firm” (web page). Accessed May 8, 2026.

National Labor Relations Board (NLRB). 2025. The National Labor Relations Board 2025 Performance and Accountability Report. Fiscal Year 2025. 

National Labor Relations Board (NLRB). 2026. “Case Search” (web page). Accessed May 8, 2026.

Pinto, Maya. 2022. How the ‘Coalition for Workforce Innovation’ Is Putting Workers’ Rights at Risk. Gig Workers Rising, National Employment Law Project, PowerSwitch Action, Service Employees International Union, and Temp Worker Justice, July 2022.

Poydock, Margaret. 2020. “The Passage of California’s Proposition 22 Would Give Digital Platform Companies a Free Pass to Misclassify Their Workers.” Working Economics Blog (Economic Policy Institute), October 22, 2020.

Rhinehart, Lynn, and Celine McNicholas. 2024. “What’s Behind the Corporate Effort to Kneecap the National Labor Relations Board?: SpaceX, Amazon, Trader Joe’s, and Starbucks Are Trying to Have the NLRB Declared Unconstitutional—After Collectively Being Charged with Hundreds of Violations of Workers’ Organizing Rights.” Working Economics Blog (Economic Policy Institute), March 7, 2024.

Shierholz, Heidi, Celine McNicholas, Margaret Poydock, and Jennifer Sherer. 2024. Workers Want Unions, but the Latest Data Point to Obstacles in Their Path: Private-Sector Unionization Rose by More Than a Quarter Million in 2023, While Unionization in State and Local Governments Fell. Economic Policy Institute, January 2024.

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. 

Key takeaways:

• Affordability is not just about prices; it’s the outcome of a race between income growth and price inflation. When income growth is slower than price inflation, affordability worsens. When income growth is faster, affordability improves.
• Focusing just on prices is bad for understanding how the economy works and how it has performed in the recent past, and it leads to an overly restrictive policy menu for improving families’ affordability.
• Policy can more reliably address income growth for typical families. This growth has been stunted for decades by the rise of inequality. Closing this gap by ensuring more equitable distribution of future growth is the strongest tool we have for improving affordability.

Affordability has been the policy buzzword of recent years. Much of the affordability discourse—both among policymakers and the public—has focused near-exclusively on prices as the big affordability problem. But affordability is not a problem of high prices, instead it’s the outcome of a race between incomes and prices. And the reason typical families have faced an affordability crunch in recent decades is not because prices have grown exceptionally fast, it’s because incomes for the vast majority have grown too slowly. This income growth has been suppressed mostly by rising inequality that has put a growing wedge between overall economic growth and the income growth of typical families.

Getting the drivers of affordability right is important—it’s not just quibbling. If you only examine price growth and try to infer what has happened to affordability over periods of economic history, you’ll usually get the story wrong. And if policymakers only look at how to change the trajectory of prices while ignoring what they can do to change the trajectory of incomes, they will be far less effective in providing useful relief to U.S. families. There are far more ways to use policy to raise incomes in a targeted and effective way than there are to suppress price growth.

Below, we provide some more background on why analyses of affordability need to include incomes, why policymakers have much more scope to raise incomes in a useful way as opposed to pushing down prices, and why focusing just on prices can obscure whether affordability has improved or worsened.

Why do prices dominate today’s affordability debates?

In modern capitalist economies, prices rise essentially every year (though at quite different rates), but so do incomes. Determining what has happened to families’ ability to afford a decent and secure life requires looking at measures that take into account both sides of the affordability equation, such as real (inflation-adjusted) income growth. Nobody really disputes this. After all, Americans could buy a new car for $600 in the 1930s, but nobody thinks society was generally richer back then.

The narrow focus on prices in assessing one’s own economic struggles likely stems from several factors.

First, inflation was very fast in the early 2020s. Americans hadn’t experienced inflation rates that high in decades, and they didn’t like them, so prices remain front of mind for many.

Second, it is true that price changes can dominate what happens to real incomes over very short time periods (say a year or less). This recognition is why we can be so sure that the oil price shock inflicted by the U.S. bombing of Iran is going to be so damaging to U.S. families. The rise in oil prices so far this year has likely baked in at least a 1.5% increase in inflation over the next 6–12 months. In 2025, real wage growth for the large majority of workers was slower than 1.5% (which was the outcome of roughly 4% nominal wage growth minus 2.5% inflation). Given this, a sharp and unexpected 1.5% jump in prices will likely erase any prospective real wage gains for workers in 2026.

Finally, it has been noted that many Americans see wage gains as something they accomplished themselves through hard work, while prices are out of their immediate control. Inflation is hence seen as damage done to them and something they need relief from. But this is mostly wrong—policy choices impact wage growth at least as much as inflation, and the most effective policy relief for living standards will come through measures that raise wages, not restrain prices.

Policy can target incomes more effectively and precisely than prices

One person’s income is another person’s cost, which means prices are a bundle of different stakeholders’ incomes. The bill you pay at the grocery store must cover payments the store makes to its shareholders, the salary of the CEO and managers, the wages of cashiers, and the cost of buying food from producers. We don’t want all these incomes to be forced down. Given extreme levels of inequality in the U.S., we would likely be fine with lower CEO pay and payments to shareholders, but we would want wages of cashiers and many in the food production supply chain to rise. Efforts to simply clamp down on this price will have uncertain effects on incomes.

In the jargon of economists, focusing on prices is sector-based policy but to genuinely improve affordability we need factor-based policies, where factors of production like capital, rank-and-file workers, and corporate management can be specifically targeted by policies that aim to raise or restrain their incomes.

Fortunately, there are many good policy options for targeted affordability policy specifically toward low- and middle-income families. Incomes for these families—and for anybody without dynastic wealth—are dominated by wages and public benefits. We talk about each of these in turn below.

Boosting public benefits is affordability policy

Public benefits are entirely under policymakers’ control. If policymakers really cared about the affordability of groceries or health care or energy, they could boost benefits for food stamps, Medicaid, and the low-income heating energy assistance program. These programs currently deliver needed assistance to tens of millions of families to make life more affordable—and they do this with vanishingly small administrative costs, meaning they are highly efficient. Yet all of these programs are slated for steep cuts in the coming decade due to the Republican tax and spending megabill passed in 2025. This bill will inflict large damage to the most vulnerable families’ ability to afford decent and secure lives.

Further, Congress and the Trump administration chose to not extend the Biden administration’s more-generous subsidies for people to buy health insurance through the marketplace exchanges of the Affordable Care Act. The failure to extend these subsidies—even after a full federal government shutdown engineered by congressional Democrats aimed at prioritizing this issue—means that average out-of-pocket costs will double for those buying insurance in the exchanges.

Besides just reversing these cuts, making the U.S. welfare state more robust could also greatly boost the affordability of a decent life. Things like making health coverage more universal with lower out-of-pocket costs, reforming unemployment insurance to make it more protective, and providing all families with children a generous universal child allowance could dramatically improve affordability.

Policy can boost affordability through higher wages as well

The link between policy changes and wage growth is slightly less direct than for public benefits, but it remains very strong. Capitalist labor markets are inherently tilted toward employers and against workers. The only periods of history that have seen strong and equal rates of wage growth across the workforce have been periods where policy supported institutions that boosted workers’ leverage with employers.

The 30 years after World War II saw the creation of policies and institutions that successfully spread the gains from rising productivity equitably among workers up and down the wage distribution, with low- and middle-wage workers seeing growth rates as fast as high-wage workers. This equitable distribution of wage growth was a crucial way that income growth more broadly was kept equitable in this period.

Since 1979, however, these institutions have been steadily attacked and weakened with no new institutions being stood up to take their place in ensuring an equitable distribution of economic growth. The result has been that wages and incomes of typical families have lagged far behind average income and wage growth (or productivity). The wedge between income growth experienced by the vast majority of families and average growth is simply income being generated in the economy that is not helping typical families’ affordability struggles. Instead, it is income being funneled reliably away to the top.

There’s no reason that the institutions that equalized wage growth cannot be built back up and modernized.

The federal minimum wage is the most obvious policy institution for raising wages at the low end of the labor market. Raising the federal minimum wage from its current shamefully low $7.25 would directly boost affordability for tens of millions of workers. In the middle of the wage distribution, unions have proven to be the institution that has historically counteracted employer power and given typical workers increased leverage. However, unions are in a far weaker position today relative to their high points because of intentional policy choices—specifically because policymakers failed to act to curb employers’ growing hostility (and often their illegal activities) toward union organizing. If stronger policy boosted union density, unions would raise wages for both members and non-members alike.

Low- and middle-wage workers also benefit enormously from a determined effort to keep unemployment low for extended periods of time. In recent decades, policymakers have tolerated excess unemployment to keep inflation in check, but this is far too costly a strategy to keep potential inflation in check. Besides locking out millions of willing workers from job opportunities, long periods of excess unemployment were periods when real (inflation-adjusted) wage growth became literally stagnant.

Policymakers often seem skeptical of the effectiveness of these wage-boosting policies, arguing that the effects are too indirect and will take too long to provide benefits to workers. It’s true that efforts to boost unionization and sustain full employment will take some time to push up wages. But they will do this reliably. Further, many policies advanced in the name of reducing prices would also take a long time to come to fruition. For example, calls to tighten antitrust restrictions against corporate mergers and to break up established monopolies often have lots of merit. However, they are not policies that happen instantly and have purely predictable effects.

Focusing too hard on prices can obscure when affordability is actually improving

Finally, one key reason to broaden the affordability debate beyond prices is simply to make sure the public and policymakers can correctly identify periods of improvement or degradation of affordability. As an example of how focusing only on prices can lead to an incorrect diagnosis of affordability trends, take the example of two five-year stretches in recent economic history, both measured from a business cycle peak and going five years forward from there: In the years between 2007 and 2012, annual inflation averaged 1.8% and peaked at 5.5%, while between 2019 and 2024, inflation averaged 4.2% and peaked at 9%. Based on price growth alone, one would expect affordability to have eroded more rapidly in that second period, and indeed the popular narrative is that the early 2020s inflation was particularly destructive for affordability.

But between 2007 and 2012, the nation’s unemployment rate averaged 8.3%, while it averaged less than 5% between 2019 and 2024. After 2007, it took 93 months to re-attain the pre-recession unemployment rate, while it took just 29 months after the 2019 business cycle peak. In short, the labor market was far stronger in the second period.

And when it comes to real (inflation-adjusted) wage growth, the second period—largely because of its lower unemployment—saw far better outcomes than the first. In the 2019–2024 period, inflation-adjusted wages for low-wage workers (those at the 10th percentile) and the median worker rose by a cumulative 15.3% and 5.8%, respectively. In short, contrary to most conventional wisdom, affordability improved in this time. Between 2007 and 2012, real wages outright fell for both low-wage and median workers. Even with very slow inflation, affordability was demonstrably worse in that earlier period.

Figure A

More recently, inflation averaged slightly lower in 2025 (2.5%) than 2024 (2.9%). Yet for many workers—and particularly low-wage workers—2025 saw weaker (or even negative) real wage growth. This is largely due to some slight cooling in the labor market as unemployment rose from 4.0% to 4.4% over the course of 2025. Hence, even as inflation decelerated, the cooling labor market led to an even faster deceleration in nominal wages, which meant that affordability worsened for many workers.

Figure B

Reducing inequality is the key to improving affordability

Because many policymakers believe that affordability concerns are a new problem caused by inflation of recent years, they are now on a frenzied search for new and creative solutions to this price problem. But because the real affordability problem for U.S. families did not emerge in the past few years (remember, affordability was improving in the five years before 2025) and because the genuine long-run problem of affordability was about the inequality of income and wage growth, not excess inflation, most of these new and creative solutions just won’t hit the mark.

It’s understandable why many policymakers seem frustrated with being reminded of the long-diagnosed problem of inequality and the proven remedies—such as sustained full employment, higher wage standards like minimum wages, protecting workers’ fundamental rights to organize unions and bargain collectively, and a more robust welfare state.

Some, of course, just don’t believe in some of these solutions, while many who do would argue that these proven remedies are politically unrealistic in the current moment. But because the real affordability problem is an inequality problem that requires those at the top of the income and wealth scales having to accept less growth going forward (less than the stratospheric gains they’ve gotten used to, it should be said), any genuine solution is going to seem impossible in today’s political system that is dominated by the wealthiest families and corporations. Any policy—whether old and well-tested or new and creative—that actually aims to redistribute income, wealth, and power away from where it sits today will face a wall of opposition that must be politically overcome one way or the other. There’s no “one weird trick” where you can develop a policy creative and neat enough that it will somehow fool the rich and powerful about what its end result will be. And if the end result of the new and creative policy does not threaten the prerogatives of the rich, it’s not a real solution.

Today’s affordability concerns are indeed rooted in objective facts about the material circumstances of middle- and working-class families in the United States. Precisely because of this, they deserve more serious analysis and policy responses than they have been getting. This means focusing more on incomes than prices, and it means being clear-eyed that it has been the upward redistribution of income to the top—abetted by policy decisions—that is the drag on typical families’ affordability. Until solutions address that, they’re mostly just noise.

Key takeaways:

• Income inequality has skyrocketed since 1979 because of intentional policy choices that suppressed wages for typical families to accelerate income growth at the top.
• Middle-class household incomes would be roughly $30,000 higher today if their incomes had simply kept pace with average income growth since 1979.
• Recognizing that today’s affordability problems are overwhelmingly inequality problems is the key to constructing the right policy solutions.
  • As a start, protecting workers’ right to organize unions, fostering long periods of very low unemployment, and keeping minimum wages high will help typical families claim their fair share of income growth.

When most people—including policymakers—complain about a lack of affordability, they think of prices being too high. But affordability is the outcome of a race between prices and incomes. After all, goods and services were a lot cheaper 90 years ago during the Great Depression, but we all know that nearly everybody is richer today than their peers back then. Bringing incomes into the affordability picture makes for better understanding and better policy.

New Congressional Budget Office (CBO) data show that rising income inequality is the main reason that affordability feels out of reach for too many U.S. families. For more than four decades, most of the income growth in the U.S. economy has been funneled to those at the very top, leaving typical families with far less than their proportionate share of the economy’s gains. If middle-class household incomes had simply kept pace with average income growth since 1979, their pay would be roughly $30,000 higher today. If we account for taxes and government transfers, incomes would still be $19,000 higher today for these middle-class households. Think of this gap as an “inequality tax”: the amount that rising inequality has cost the typical U.S. family. Life would be much more affordable for these families today if they hadn’t been hit by this inequality tax.

This inequality is not the result of competitive markets fairly rewarding people’s skills and hard work. Instead, it resulted from an intentional policy campaign of wage suppression. Labor markets in capitalist economies are inherently tilted toward employers. Fair pay and broadly shared prosperity only materialize when policy affirmatively aims to correct this power imbalance. This can happen—policy choices that bolstered workers’ leverage and bargaining power in labor markets kept growth fast and equal for decades following World War II, for example. But lawmakers rolled back these policies at the behest of capital owners and corporate managers.  

The latest CBO inequality data make the scale of this policy shift visible. Figure A shows the distribution of market income growth for non-elderly households by income group since 1979. We use market income to look at pre-tax, pre-transfer outcomes to assess the equality of outcomes generated by markets. We isolate non-elderly incomes because older households tend to have very low market incomes and these older households have grown as a share over time—so we don’t want any poor performance of market incomes documented here to simply be the outcome of natural population aging. Among this non-elderly group, the top 1% have captured a hugely disproportionate share of market income growth. Between 1979 and 2022, market income for the top 1% grew 277% (from $784,573 to $2.958 million) compared with just 26% growth for the middle fifth of households (from $76,359 to $96,335). This lopsided growth is the root of America’s affordability problem. Even as the economy grew and average incomes rose, typical families fell further behind those at the top who captured most of income growth.

Figure B shows the inequality tax over time, plotting actual market income for the middle fifth of households against what their income would have been if it had grown at the same rate as overall average income. By 2022, the inequality tax reached $30,676 per household, meaning middle-class families are forgoing that much income each year because of rising inequality. The gap has widened steadily since 1979, a sign that the affordability problem facing typical families is not a recent development but rather the cumulative result of decades of policies that have shifted income upward.

Because market income for middle-class families is driven predominantly by labor income, the inequality tax in Figure B reflects the consequences of decades of wage suppression. Of course, the United States has a system of taxes and means-tested transfers (safety net programs like Medicaid and food stamps, for example) that leads to post-tax and transfer income being more equal than market income in any given year. But the tax and transfer system did not ramp up in importance as market income inequality grew after 1979, and even after accounting for its effects, inequality increased significantly. Figure C shows that even when using post-tax and transfer income, the inequality tax remained substantial at $19,320 per middle fifth household in 2022.

Figure D shows who loses and who gains from rising inequality. While the inequality tax cost middle-income families $19,320 in 2022, families at the very top benefited enormously. The 96th to 99th percentiles gained about $88,000 from rising inequality, while the top 1% gained $1.1 million in 2022.

Perhaps surprisingly, the lowest quintile also slightly gained. For this group, lower taxes and higher levels of means-tested benefits counterbalanced a significant loss of market income due to inequality (their market income inequality tax would be around $4,000). The greater fiscal transfers to the bottom fifth are an under-recognized policy achievement of recent decades. It is also an achievement under constant threat, with the latest one being the large cuts to Medicaid and food stamps coming because of the Republican tax and spending bill that passed in 2025.

Figure D

U.S. families’ feeling that life is less affordable than it should be is grounded in objective realities about how the economy has failed them. And it’s understandable why so many of these families think about prices, which they see as the final barrier between them and being able to obtain what they need for a good life, whether the price is for a gallon of gas or a loaf of bread or a monthly health insurance premium.

But the forces causing this affordability crunch are far larger than any given set of prices. Instead, they are mostly the forces that led to rising income inequality by intentionally suppressing the power of workers in labor markets. This wage suppression meant that middle-class income growth was never going to outpace inflation consistently enough to ensure steadily improving economic security.

In short, today’s affordability problems are overwhelmingly inequality problems. Recognizing this fact is the key to constructing the right policy solutions. As a start, protecting workers’ right to organize unions, fostering long periods of very low unemployment, and keeping minimum wages high will help typical families claim their fair share of income growth.

Key takeaways:

• Recent Democratic proposals to exempt broad swaths of the middle class from federal income taxes accept a damaging frame of taxes as a pure drain on affordability.
• But taxes aren’t a drain on affordability; they fund the public services and social insurance programs that make a decent life possible for middle-class families.
• Progressive taxes on the ultrarich and corporations are essential and should be the immediate priority, but they cannot sustain the public sector alone, let alone expand it in ways needed.
• Middle-class tax rates have fallen by a third since 1979, yet economic anxiety remains high. Tax-cutting has failed because it has left the private-sector drivers of inequality untouched and starved public services.

For decades, anti-tax politicians have tried to smuggle in large tax cuts for the ultrarich and corporations by loudly offering tax cut crumbs to the middle class. Key to this effort has been framing taxes as a pure drain on typical families’ ability to afford a secure economic life. Any success in this dishonest campaign to foster anti-tax sentiment is a disaster for working people—and that’s why some recent tax policy ideas from Democrats and the rhetoric around them are so deflating.

Two things are true about taxes in the United States. First, taxes on the richest families and corporations are far too low. Second, it is broad-based taxes on the middle-class that are the foundation of a functioning public sector and a decent society.

Progressive taxes on the ultrarich and corporations are mostly needed to reduce the potential gains to the rich and powerful from rigging the rules of markets. When the powerful rig these rules and hugely disproportionate shares of income concentrate at the top—like in the United States today—progressive taxes can also raise significant revenue.

But if sharply progressive taxes succeed in reducing the incentive for rigging the rules of markets and if other policies help lead to more broadly shared income growth in the country, this means that progressive taxes will raise a lot less revenue over time.

To be clear, this would be a victory for a better society. For example, the purpose of a carbon tax is to lower greenhouse gas emissions and if it’s highly successful, it will by definition stop raising much revenue. Progressive taxes aimed at reducing inequality will see the revenue they raise start to decline when they are their most successful. Right now, we do have deep inequality in the U.S. and progressive taxes will raise a lot of money—but we shouldn’t make the public sector’s resources dependent on this remaining true forever.

But more importantly, taxing only the very rich has never been the primary foundation of public-sector resources and can’t be going forward. The revenue needed to support programs that provide social insurance and income support (Social Security and Medicaid, for example), as well as public investment and services like highways, transit, and public education requires broad-based taxation. Without Social Security providing secure retirement, Medicare, Medicaid, and the Affordable Care Act providing access to health care, and public schools providing universal education, a decent life for the middle-class would be entirely unaffordable. And without middle-class taxes supporting all of these things, they would collapse.

If typical Americans lose faith that paying broad-based taxes to support public services and investments is a good deal, it will be a disaster for their ability to afford a decent life. Sadly, some recent Democratic proposals capitulate to this view of taxes as a “pure burden” rather than an investment in the country and its people.

Senators Chris Van Hollen (D-MD) and Cory Booker (D-NJ) have both floated ideas that would draw a line below which nobody would pay any federal income tax (Van Hollen’s line is $92,000 for a married couple while Booker’s is $75,000). Both proposals pay for this with tax increases on the rich. If these tax increases on the rich were standalone pieces of legislation, they’d be excellent. But they are instead paired with tax benefits that mostly miss the bottom of the income distribution. In Booker’s proposal, the gains from the higher standard deduction that zeroes out taxes for many are actually largest for families between the 60th and 80th percentiles, and gains persist on average through the 99th percentile. Van Hollen’s bill phases out the “alternative maximum tax” that zeroes out taxes for many, and the biggest gains hit in the middle of the income distribution, but the proposal still provides gains on average for families between the 60th and 80th percentiles.

Both proposals clearly aim to address the affordability challenge that politicians have seized on, but in doing so both frame taxes as a pure drain on affordability, with Booker even calling his the “Keep Your Pay Act.” But taxes aren’t a drain on affordability. They provide the resources needed to run the public sector, and the public sector in turn does a great deal to make life more secure and more affordable over people’s lifetimes.

Social Security and Medicare, for example, both rely on payroll taxes on workers’ wages. But they also provide income for these workers in retirement. Instead of draining affordability, these programs smooth income over the lifecycle to ensure working families can afford a decent life even when they can no longer work. Food stamps and Medicaid are financed by taxes and provide benefits to people who otherwise would not be able to afford the most basic necessities: food and health care. The same people who receive Medicaid and food stamps in one era of their lives will contribute to them through taxes in other periods when they have found steady work. Again, the taxes collected are recycled back into families’ incomes in ways that minimize suffering and severe affordability crises throughout their lives.

State and local taxes—often borne quite heavily by the broad middle class—pay for public education. This education—both K–12 and higher education—is incredibly valuable and necessary for anyone operating in modern economies. Without the taxes to support education, families would have to dig into their own pockets to pay for private schooling, and it would be delivered less efficiently and much less equitably.

Other taxes finance infrastructure and other key public goods and services, without which life would be harder and more expensive for most families.

Cutting taxes even fails on the crass political grounds of buying voters’ short-term goodwill. It’s often underrecognized (mostly, again, because of conservative campaigns to hide this fact), but taxes for the middle class have been cut a lot in recent decades. Figure A below shows the percentage point change in tax rates of households at different parts of the income distribution between 1979–2019. We stop at 2019 to compare equivalent points of the business cycle. Tax rates tend to fall sharply during recessions, which can obscure the full extent of legislative changes to tax rates. Further, cutting taxes temporarily during recessions can make some sense—tax cuts are one form of fiscal stimulus that can be used to fight recessions (unless these tax cuts are quite well-targeted on low- and moderate-income families, they tend to be less efficient stimulus than spending measures).

Figure A

The largest tax cuts have gone to the bottom fifth of households—a key policy victory of recent decades. The expansion of refundable tax credits like the Earned Income Tax Credit and the Child Tax Credit—credits that are paid directly to lower-income families even if their amount is greater than the families’ tax liability—have essentially made the problem of taxing families into poverty almost nonexistent. These tax cuts should clearly be kept. But all households, not just those with very low incomes, have seen sizable tax cuts. Tax rates for households in the middle of the income distribution have been cut by a third since 1979.

And yet, does anybody feel like this tax-cutting has led to most U.S. households feeling great about their place in the economy and their prospects for affording a decent life today? Do these voters express warm feelings about the policymakers from both parties who provided these middle-class tax cuts?

The tax-cutting strategy has failed to make these households happy for two reasons. First, it leaves the private-sector drivers of inequality untouched, and as governments have collected less in taxes, employers and corporations have contributed less to middle-class families’ wages. Second, lower taxes have starved public-sector capacity and led to a degradation of public services. Strangely, the newly fashionable “abundance” movement often frames this degradation as a problem of public-sector excesses, but it’s clearly driven by disinvestment. In short, middle-class families value public services and the decades-long campaign to cut taxes has harmed the ability to provide them. The lessons for today’s tax debates should be clear.

The failure of tax-cutting to foster economic security and happiness is not all that surprising for scholars of U.S. attitudes toward taxes, who argue that Americans are not universally anti-tax. Instead, Americans view paying taxes as a patriotic good and a moral obligation. But they are angry about paying their taxes when they think others are shirking their part of the social contract, particularly when they think the richest people and corporations aren’t paying their fair share.

Because we are starting with such high levels of inequality and because of this public cynicism about the rich ever being forced to pay their fair share, the first priority—by far—for policymakers today should be to enact significant stand-alone tax increases on the ultrarich and corporations. The revenue raised solely from the ultrarich could close today’s fiscal gap, the difference between today’s budget deficits and what is needed to put them on a sustainable path going forward. And this act would convince the rest of Americans that the ultrarich are not always prioritized in policymaking and would make future debates about the costs and benefits of higher taxes for higher levels of public goods much healthier.

But we can’t run a decent society based on just taxing the rich and telling everybody else that taxes are an unfair drain. Oliver Wendell Holmes famously said that taxes are the price you pay for civilization. But if the taxes are paid only by the rich, we will get the civilization they want. That doesn’t seem good enough to me.