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.

The Trump administration has consistently prioritized aggressive and arbitrary immigration enforcement, with the ultimate goal of deporting 1 million people every year of his term—regardless of their contributions to their communities and the U.S. economy. While the Department of Homeland Security’s pace currently falls short, increased enforcement would curtail business operations and reduce employer demand for both immigrant and U.S.-born workers. Over four years, 1 million annual deportations could cause total employment in the United States to fall by 5.9 million jobs, with particularly severe losses in construction and child care industries.

The direct care sector is also highly vulnerable to these enforcement actions. Amanda Kreider and Rachel Werner’s recent research indicates that job losses will significantly affect workers who provide long-term care in home- and community-based settings. The direct care sector—which includes home health aides, personal care aides, orderlies, psychiatric aides, and some nursing assistants—relies heavily on immigrant labor. Immigrants constitute nearly 30% of the direct care workforce, compared with 20% of overall employment. Among home health aides who assist with daily living and healthcare tasks, four in 10 workers are immigrants.

Kreider and Werner found that previous increases in immigration enforcement caused the direct care sector to shrink. If these patterns hold under the current enforcement regime, four million deportations over four years could cause direct care employment to fall by 394,000 (see Figure A).

Figure A

The majority of this employment decline—274,000 jobs—will result from the loss of immigrant workers. However, in addition to removing a supply of labor, deportations also make the labor market more precarious for immigrant workers. When immigrants face heightened risk of arrest, detention, or deportation, their ability to change jobs becomes severely constrained. With reduced labor market leverage, employers can worsen working conditions and suppress wages for all workers in the sector, not just those directly affected by deportations.

Contrary to the misconception that deportations will increase job opportunities for U.S-born workers, existing research consistently demonstrates that increased immigration enforcement reduces the employment for both immigrant and U.S.-born workers. Deteriorating pay and conditions for direct care workers would make U.S.-born workers unlikely to step in to replace the shortfall of immigrant workers, consistent with what studies have found when immigration enforcement decreased the size of the construction and child care sectors. In direct care, about 30% of the employment decline—the equivalent of around 120,000 jobs—will affect U.S.-born workers.

While employment reductions will be widespread, they will hit certain states particularly hard due to the geographic concentration of noncitizen immigrants in the direct care sector (see Table 1). New York faces especially severe challenges. Immigrants comprise two-thirds of the state’s direct care workforce, and more than one-third of all noncitizens working in direct care nationwide live in New York. If the Trump administration achieves its deportation goals, New York’s direct care sector could shrink by 45%.

Table 1

These large employment losses would translate directly into reduced availability of direct care services. Kreider and Werner found that past escalations of immigration enforcement led to substantial increases in the number of older adults living without any help at home. Among the Medicaid population, formal nonfamily caregiving declined while family-based caregiving increased, reflecting the contraction of the formal direct care sector.

This shift from formal to family-based care suggests that job losses in the direct care sector will have large spillover effects across the economy, greatly increasing their potential harm to even U.S.-born workers. As direct care supply becomes constrained due to deportations, some family members may need to leave their jobs or reduce their work hours to assume new caretaking responsibilities. Indeed, other research has shown that increases in immigration enforcement caused U.S.-born mothers to work fewer hours due to declining availability of household services like cleaning and child care. Family members may well be forced to choose between their careers and caring for aging and disabled relatives.

The Trump administration’s deportation agenda threatens to trigger a cascading crisis in senior and disability care that will harm families across the economic spectrum. Even in the absence of deportations, caretaking needs will accelerate as the older population grows tremendously, especially in the next five years. If the direct care workforce contracts by nearly 400,000 workers due to deportations, millions of older adults and people with disabilities will be left without the professional assistance they need to remain safely in their homes. Rather than creating jobs for U.S.-born workers as proponents claim, mass deportations eliminate employment opportunities for citizens and immigrants alike while dismantling a care infrastructure that seniors, people with disabilities, and families depend on.

Given Congress’s inability to pass legislation protecting workers’ wages and health and safety, federal regulations are critical to ensuring that workers have meaningful protections. However, Trump and his Department of Labor (DOL) recently announced a massive deregulation effort, robbing U.S. workers of dozens upon dozens of rules that protect them from being forced to risk illness and injury on the job and ensure that they are paid for their labor.

Federal regulations are written by agencies to carry out the specifics of laws passed by Congress. Since laws can remain unchanged for decades, regulations make sure that laws are implemented and administered effectively. Since returning to office, Trump has unleashed a deregulatory agenda that requires agencies to identify 10 regulations to repeal in order to enact one new regulation. This framework perpetuates the misguided belief that regulations are burdensome for employers and harm economic growth. On the contrary, the long-term benefits of regulations consistently outweigh the costs—and research shows their impact on jobs is either neutral or modestly positive.

More than half of the regulations Trump’s DOL proposes to rescind are designed to ensure that workers have safe workplaces, such as requiring employers to limit workers’ exposure to a variety of harmful substances like benzene, asbestos, lead, and cotton dust.

Other rules being eliminated protect workers’ wages. These include the 2013 home health care worker rule, which gave in-home care workers the right to earn the minimum wage and overtime pay for the first time. DOL also wants to roll back the 2024 farmworker protections rule, which would have helped farmworkers advocate for better working conditions and required other protections for this extremely vulnerable group of low-wage workers. Trump’s DOL also withdrew a proposed rule that would have eliminated the subminimum wage for workers with disabilities, effectively keeping many people with disabilities unable to receive fair pay for their work.

Labor Secretary Lori Chavez-DeRemer framed this massive deregulatory effort as eliminating unnecessary regulations that are burdensome and obsolete. To be clear, there is nothing obsolete about rules that protect workers from exposure to harmful substances. These substances have not become less harmful over time—they still sicken, injure, and kill workers each year. In 2023, more than 800 workers died from exposure to harmful substances and environments. There is also nothing obsolete about rules that ensure workers are paid for their labor, especially when employers are estimated to steal $15 billion from workers’ paychecks each year.

Despite Trump’s rhetoric and the Secretary of Labor’s confirmation promises, Trump and Chavez-DeRemer are clearly advancing an anti-worker agenda. Trump’s DOL proposes to eliminate regulations that protect workers’ health and safety and make sure workers are fairly compensated for their work. This is not an agenda that serves workers—in fact, the repeal of these rules is a gift to corporate bad actors who resist complying with the rules.  

Foreword • Authors • Acknowledgements

A joint project of Center for American Progress, Center for Popular Democracy, Economic Policy Institute, Groundwork Collaborative, National Employment Law Project, National Women’s Law Center, and Washington Center for Equitable Growth.

Foreword

Heeding the voices of workers rising up to fix a system that failed us

If nothing else, COVID-19 has taught us that if we don’t make bold changes to our nation’s economic and social foundation now, we could squander our best chance to save our economy from tanking in the next decade. We workers are that foundation, and this crisis has revealed fissures in that foundation that have been expanding for the past half-century.

I’m a Black Native American woman who just turned 60. I’m also an award-winning former broadcast and print journalist. My 40-year-plus career path has been obstructed by workplace abuse, systemic discrimination, racial and gender bias, wage suppression, and now ageism. I was already way behind the eight ball when the pandemic struck. And now I am unemployed.

Latrish Oseko and Dora Whitfield have faced similar obstacles. I learned Latrish’s and Dora’s stories through our leadership in a movement of unemployed workers. While Dora, Latrish and I established our careers in different industries and live in different states, we have faced the same insidious race and gender biases in our workplaces—and the same challenges in rejoining the ranks of the employed. Our unemployment experiences support the problems flagged in this report: that people of color, especially African Americans, represent a historically disproportionately large percentage of COVID-unemployed workers, and are least likely to get the benefits we need to get back on our feet. We’re the last hired, first fired, and the last rehired.

At the end of 2019, I was freelancing and about to start training for my second year with the U.S. Census Bureau when I clinched a remote long-term freelance opportunity providing content strategy to the University of Maryland. It was the next break I needed.

Then COVID-19 hit. I knew the CARES (Coronavirus Aid, Relief, and Economic Security) Act provided Pandemic Unemployment Assistance (PUA), the federal program bringing jobless aid to workers ineligible for traditional UI. For the first time, the UI system offered benefits for nontraditional “1099 workers” like me. Without this temporary assistance, I, and millions of other workers, would have been without any financial support this past year. That’s why this report recommends that all workers, including contract, gig, and self-employed workers, be permanently eligible for some form of UI benefits.

Having been on Medicaid and food stamps in the past, I know how easy it is to answer an application question wrongly and be disqualified. My PUA application was approved and processed by early May. But without warning, the benefits stopped. I searched Facebook for groups that were sharing UI application information. I found one: “Georgia Unemployment Issues—COVID-19.” Our members included newly jobless people from all sectors. We had unintentionally joined a burgeoning national movement of people engaged in mutual aid and activism to improve a dysfunctional UI system.

Things got much, much worse after the additional $600 weekly benefits provided by the federal Pandemic Unemployment Compensation (PUC) program ended July 31. More and more of our Facebook group members shared their heartbreaking stories of pending evictions, food insecurity, lost medical insurance, repossessed cars, and declining mental health. While the PUC program was extended at a lower weekly benefit level of $300 after a lapse of several months, it didn’t provide enough help for people falling further and further behind.

Coping with inadequate and unreliable benefits even in normal times is a familiar story for many workers. Louisiana, for example, provides notoriously stingy benefits. States with higher Black populations tend to have shorter benefit durations and less generous benefits. In these states, which tend to be in the South, many jobs are low-wage jobs and low benefits serve to force Black workers into those low-paying jobs. It’s one reason why they pay benefits that are way below the cost of living. It’s a big reason why we need federal standards.

Dora Whitfield had worked for 30 years in the New Orleans hospitality industry. She’s not well-to-do, but as a union member she enjoyed a stable income—enough to purchase her first home in 2014. Then the pandemic pulled the rug from under her. Her employer reduced her hours to part time until she was furloughed this past September.

Dora, a member of Step Up Louisiana, a community organization that organizes for economic and education justice, had never been on unemployment. She received the state’s maximum UI benefit of $247 a week, $221 after taxes.

“$247 a week?! Do you know I make that much money in tips when I work on a weekend? That’s on top of my regular paycheck,” she says. “When you have $221, it’s like you’re playing Russian Roulette. You have one bullet in your gun. And you have your light bill here, your water bill here, your car insurance, and then you might have your house note. So, wherever you shoot that bullet, that’s when you have to decide what bill you’re gonna pay.”

In Delaware, Latrish Oseko’s company began reducing its contract workers’ hours to part time in March. By May, all contractors were laid off, leaving Latrish, her live-in boyfriend, and their 4-year-old daughter solely dependent on his income. She lost her job the same month her landlord decided to evict her family. Despite their search for new housing, they hit the same wall millions of jobless Americans have encountered.

“We applied everywhere but we couldn’t get an apartment to save our lives because no one would accept us on my boyfriend’s measly income at the college and they would not accept my unemployment as income,” she explains.

They landed in Motel 6, spending more than $2,500 a month. The weekly PUC benefits she had saved kept a roof over their heads, but her savings dwindled quickly. Latrish’s boyfriend worked a job that put him at high risk for infection. His exposure forced the family to quarantine six times. Eventually, he contracted the virus. Latrish withdrew her daughter from day care, further frustrating her job search and causing another hit to the family’s income.

“We’re in a no-win situation,” she adds, noting that her current benefit amounts don’t cover her cost of living.

Dora and Latrish’s stories show that the unemployment system can be a literal lifeline, as when it kept Latrish’s family from living on the street. But far more often, UI fails working people—especially Black women like us. When UI is too stingy or unavailable due to narrow eligibility, millions of working people are either struggling to just survive, or forced to take work that is underpaid, unsafe, and offers no stability.

Employer lobby groups have persuaded governors around the country to withhold federal UI benefits because they refuse to raise wages or job quality to attract job seekers. Instead, they prop up a system that coerces us into jobs that pay below the cost of living and require working multiple jobs, consuming all our waking hours with no guarantee of steady hours or income. Without a federal right to unemployment benefits and the other recommendations in this report, many workers in this country will be trapped in a cycle of poverty and economic stress that, quite literally, is killing them.

In short, good jobs—jobs that allow us to live in dignity and security—aren’t available to a large swath of the American workforce. Before the pandemic, good paying jobs were hard to find after the Great Recession. Now, as post-pandemic recovery looms on the horizon, a weary workforce can’t afford more cuts in wages and stability.

Yet, we, the COVID-unemployed, have been accused of choosing a UI-benefits-based lifestyle instead of rejoining the workforce. Our answer to that narrative is a resounding “no!” We simply want work that pays wages and salaries that cover the rising cost of living. We want to grow in our aptitude, talents, and skills, and use them as they should be used—to make a reasonable living. Instead, we are being asked to accept a form of financial abuse by accepting work that traps us in a cycle of poverty and stagnation. So, we are pushing back. It’s that simple.

Our firsthand experiences underscore why this report proposes federal benefits standards to make sure benefits are enough to survive on. Our experiences show why we need a dependent allowance of $35 per week. They demonstrate why we need automatic triggers to extend benefits up to 99 weeks when the true unemployment rate increases.

These critical reforms would not only save vulnerable households but also protect the economy. Finally, the universality of our experiences underscores the need to finance unemployment benefits with federal taxes, not state ones, and ensure our access to benefits is not subject to the whim of state officials.

The obstacles we’ve faced are why we got involved in organizing for economic and racial justice. The proposals here were developed in collaboration and consultation with Latrish, Dora, me, and other workers like us. We need a better foundation for our economy. A reformed unemployment system with federal standards that ensure equity, efficiency, and stability can reinforce our weakened foundation. A well-crafted policy structure of fairness that is baked into the framework will provide a sturdy rebar system that benefits and empowers all workers.

—Sharon Shelton Corpening, Roswell, Georgia

Authors

Organizations noted for identification purposes only

Josh Bivens (Economic Policy Institute), Melissa Boteach (National Women’s Law Center), Rachel Deutsch (Center for Popular Democracy), Francisco Díez (Center for Popular Democracy), Rebecca Dixon (National Employment Law Project), Brian Galle (Georgetown University Law Center), Alix Gould-Werth (Washington Center for Equitable Growth), Nicole Marquez (National Employment Law Project), Lily Roberts (Center for American Progress), Heidi Shierholz (Economic Policy Institute), William Spriggs (AFL-CIO and Howard University), and Andrew Stettner (The Century Foundation)

Acknowledgments

The contributors and sponsoring organizations also appreciate the engagement and constructive feedback from key stakeholder organizations, including the American Federation of State, County and Municipal Employees (AFSCME), Service Employees International Union (SEIU), United Food and Commercial Workers International Union (UFCW), and UNITE HERE.

Lora Engdahl’s tireless and insightful editing made this report possible. We are grateful for all of the work Amee Chew provided throughout the drafting of this report. Doug Steiger provided outstanding support in drafting and editing. Elizabeth Pancotti and Kitty Richards were invaluable before departing for government service. Key legal analysis and feedback on the challenges facing workers came from Julia Simon-Mishel. Don Rhodes and Cornelius Cornish Jr. provided critical fact-checking assistance; Barbara Karni and Teresa Kroeger provided editing assistance; John Carlo Mandapat, Daniel Perez, and Eric Shansby assisted with layout and design; and Colleen O’Neill provided a final proofread.

The members and leaders of Unemployed Action, a project of the Center for Popular Democracy, provided invaluable influence, input, and insights on their experience of the UI system. This report would not have been possible without them, and many of the recommendations wouldn’t have been possible without their perspectives and organizing. We’re grateful for the aid provided by Gus Leinbach and Arpan Patel in helping facilitate this process. This would not have been possible without the workers and worker-leaders who took part in the series of workshops that took place over two months and provided critical influence, feedback, and input on the policy to the authors. Similarly, staff and members of the following organizations were heavily involved in that process, including Benjamin Zucker, Gabrielle Bolden-Shaw, and Caleb Holmes of Step Up Louisiana; Lalo Montoya of Make the Road Nevada; Emily Dhatt and Sage Wilson of Working Washington; Claire Galpern of One PA; Nicole Fears-Washington, Klaire Gumbs, and Eric Robertson of the New Georgia Project; and Kalia Johnson and Molly Shack of the Ohio Organizing Collaborative.

We also thank the many individuals who took time to provide us with feedback on the full report or on individual chapters (organizations noted for identification purposes only):

David E. Balducchi (U.S. Department of Labor, retired), Lauren Bauer (Brookings Institution), Alex Camardelle (Georgia Budget and Policy Institute), Jaya Chatterjee (SEIU), Cecelie Counts (AFL-CIO), Martha Coven (Princeton School of Public and International Affairs), Mia Dell (SEIU), Indivar Dutta-Gupta (Georgetown Center on Poverty and Inequality), Wendy Edelberg (Brookings Institution), Kathryn Anne Edwards (Rand Corp.), Michael Graetz (Columbia Law School), Kali Grant (Georgetown Center on Poverty and Inequality), Steve Gray (National Employment Law Project), Monica Halas (Greater Boston Legal Services), Sarah David Heydemann (National Women’s Law Center), Hilary Hoynes (University of California, Berkeley), Alex Jacquez, Andrew Johnston (University of California, Merced), Sowmya Kypa (SEIU), Marty Leary (UNITE HERE), Sharit Cardenas Lopez (SEIU), Rachel Lyons (UFCW), Elaine Maag (Urban-Brookings Tax Policy Center), David Madland (Center for American Progress), Rachel Melendes (UNITE HERE), Michael Miller (U.S. Department of Labor, retired), Gwen Mills (UNITE HERE), Shaun O’Brien (AFSCME), Robert Pavosevich (U.S. Department of Labor, retired), Matt Pearce (Media Guild of the West), David Ratner (Federal Reserve), Kelly Ross (AFL-CIO), Jesse Rothstein (University of California, Berkeley), Carrie Sallgren (UNITE HERE), Zach Schiller (Policy Matters Ohio), Garrett Andrew Schneider (SEIU), Paul Schwalb (UNITE HERE), Alexa Tapia (National Employment Law Project), Wayne Vroman (Urban Institute), and George Wentworth (National Employment Law Project).

Suggested citation

Bivens, Josh, Melissa Boteach, Rachel Deutsch, Francisco Diez, Rebecca Dixon, Brian Galle, Alix Gould-Werth, Nicole Marquez, Lily Roberts, Heidi Shierholz, William Spriggs, and Andrew Stettner. 2021. Reforming Unemployment Insurance: Stabilizing a System in Crisis and Laying the Foundation for Equity. A joint report of the Center for American Progress, Center for Popular Democracy, Economic Policy Institute, Groundwork Collaborative, National Employment Law Project, National Women’s Law Center, and Washington Center for Equitable Growth. June 2021.

Key takeaways:

• Because older workers are more likely to be unemployed for long periods, have work-limiting disabilities, and live in areas of the country that were struggling even before the crisis, policies aimed at addressing these problems will especially benefit these workers.
• While infrastructure spending could help jump-start the post-pandemic recovery, policies must ensure that older workers participate in training and jobs programs related to these investments.
• Regulatory protections for front-line workers, especially older workers and others at heightened risk for contracting or suffering serious consequences from contagious diseases, need to be strengthened and updated using lessons learned from the pandemic.
• Employer-provided benefits result in spotty coverage and higher costs for older workers. The United States should catch up to other countries and provide sick leave, paid family leave, and health insurance through government programs rather than leaving these to the discretion of employers.

(See the companion blog post outlining steps needed to protect vulnerable older workers in the economic collapse caused by measures needed to combat the COVID-19 pandemic.)

Once the worst of the outbreak is over and social distancing measures are relaxed, policies to help older workers will be needed to ensure they share in the recovery.

Deficit-financed stimulus spending—needed to quickly bring the economy back to something approaching full employment—will help but not ensure broad-based prosperity. Policymakers also need to address power imbalances between employers and workers and target policies at disadvantaged workers, including unemployed older workers.

Older workers, as I discussed in my last blog post, may find it harder to get back in the job market after layoffs for a number of reasons. They may have health conditions that limit what they can do or they may feel forced to accept large pay cuts because some skills and knowledge they’ve built up aren’t transferable and may be undervalued by prospective employers. Absent policies to help these workers regain their footing, they may become “discouraged workers” who give up on the job search and retire before they’re ready to.

This post lays out a series of policies to address barriers to employment for unemployed older workers and to protect older workers from health and financial risks.

Consider policies to encourage employers to hire long-term unemployed workers

Older workers who lose their jobs in a recession are more likely to be unemployed for long periods. For example, a study of unemployed workers in the Great Recession found that only a third (34%) of adults ages 62 and older who lost jobs were reemployed within 12 months and only two-fifths (41%) were reemployed within 18 months. On average, an older worker’s chance of reemployment each month was about half that of the average worker age 25 to 34.

Policies to encourage employers to hire long-term-unemployed workers must be carefully assessed to gauge their effectiveness and minimize the negative impact on other workers. For example, enacting the Long-Term Unemployment Elimination Act would provide time-limited funding to local workforce development boards and community-based organizations to employ workers who have been unemployed for six months or more, along with providing other employment supports. Importantly, it includes provisions to discourage employers from displacing existing workers or rotating eligible workers through the program.

Currently, the Work Opportunity Tax Credit provides a targeted employment subsidy worth up to $6,000 for one year to employers who hire workers who receive government benefits, are disabled veterans, were previously incarcerated, participate in Vocational Rehabilitation programs, or are long-term unemployed.

Ensure older workers have a role in infrastructure projects

Low interest rates, unmet needs, and bipartisan support make it likely that infrastructure spending will play a role in stimulating the post-pandemic recovery. A Green New Deal and other infrastructure measures could help jump-start the economy and address the even bigger threat to human life and the global economy posed by global warming. However, we need to ensure that workers of all ages are trained and employed for this work. As a quick start, boosting transit funding, especially for buses and accessible transportation, could have multiple benefits for older workers, who are more likely than younger workers to be employed in the transportation sector. (Unless otherwise noted, all references to older workers’ employment shares are based on the author’s analysis of 2015–2017 American Community Survey microdata for workers ages 55–64.)

Another low-tech but effective way to reduce reliance on fossil fuels is rehabbing older housing to make it more energy efficient. Barriers to making these ultimately cost-saving investments, especially for low-income and elderly homeowners, include upfront costs, lack of information, and an understandable reluctance to have a work crew in one’s home. Though construction projects tend to favor younger workers, outreach efforts to overcome these barriers in communities where declining industries have left behind an aging workforce and a dilapidated housing stock could employ many lower-income older workers.

Other forms of job training and public investment, such as universal pre-K, should also take older workers into account. Many older workers currently work as teachers’ aides, and more could be trained to work in preschools.

Better target aid to hard-hit regions

More effective and better-targeted regional economic policies would also benefit older workers, who are more likely to be employed in declining industries and live in economically depressed areas. Unemployed older workers are often less able to relocate to find work because of family commitments and community ties. Homeowners in depressed areas may also have difficulty relocating if the value of their home has declined relative to the outstanding balance on their mortgage. Eligibility for some benefits, such as extended Supplemental Nutrition Assistance Program (SNAP) and unemployment insurance benefits, is partly based on regional economic indicators. These benefits are well targeted and should be expanded. However, the Opportunity Zones tax break enacted in 2017 with the stated goal of encouraging investment in low-income areas mainly benefits wealthy investors and contributes to gentrification.

Expand EITC benefits for workers without dependent children

Another overdue policy reform that could help older workers in the recovery is expanding eligibility for the Earned Income Tax Credit. The EITC in its current form depresses the earnings of older workers competing for the same jobs as workers with dependent children who are the main beneficiaries of the program, because supplementing the incomes of some low-paid workers allows their employers to pay lower wages. We should expand EITC eligibility and benefits for workers without dependent children, while increasing the minimum wage to offset the implicit subsidy the EITC provides to low-wage employers.

We should also consider expanding and promoting the Senior Community Service Employment Program, which provides temporary subsidies for training and employing low-income older workers. This program appears to have at least modest success at promoting longer-term employment.

Institute fair-hiring policies for formerly incarcerated people reentering the labor market

Among those workers who will face the greatest hurdles to employment are formerly incarcerated older Americans. Some states and counties have suspended bail and released prisoners early in response to the pandemic, and U.S. Attorney General William Barr has announced that more federal prisoners will be eligible for home confinement. These initiatives urgently need to be replicated around the country, especially for older prisoners who pose no threat to society. However, formerly incarcerated people reentering the labor market, whether on schedule or early, are doing so at the worst possible time. In addition to ensuring that economic stimulus measures are timely and sufficient to restore a tight labor market, we should promote fair hiring for people with arrest or conviction records.

Enhance use and enforcement of federal worker safety protections

The pandemic has laid bare the weakness of worker health and safety protections in the United States. Under the Trump administration, the Occupational Safety and Health Administration (OSHA) has failed to use even the weak powers at its disposal, let alone instituted an Emergency Temporary Standard for Infectious Disease to protect front-line workers in the pandemic. Protections for older workers, whistleblowers, and others at heightened risk need to be strengthened and updated to take into account lessons learned from the pandemic.

Protect and expand disability insurance and other social insurance programs

The vulnerability of older workers with underlying conditions in the COVID-19 pandemic should also serve as a caution against reforms aimed at moving beneficiaries off disability rolls and into the workforce, as the Trump administration proposed in its 2021 budget (though how the administration intended to accomplish this was unclear). Reforms to move people off of disability rolls are especially ill-advised in a weak labor market. Instead, we should increase access to benefits while expanding supports for workers with disabilities who are able to remain in the workforce.

Strict eligibility standards and a lengthy and complicated application and appeals process make it difficult for workers to access Social Security Disability Insurance benefits. Though SSDI applications are likely to rise sharply as more workers with disabilities lose their jobs, acceptance rates typically drop in recessions due to an increase in marginal applicants. Applicants with less education, language barriers, and transportation challenges are especially disadvantaged by the complicated application and appeals process. These barriers to access have only increased in recent years, contributing to a sharp drop in take-up over the past decade.

Many older workers with disabilities don’t even bother trying to apply for SSDI and simply apply for reduced Social Security retirement benefits at the early eligibility age of 62. This is disadvantageous, since disability benefits at any age are based on retirement benefits at the normal retirement age (currently 66), not benefits reduced for early retirement. Importantly, disabled workers who may be eligible for either type of benefit and can’t afford to wait to see if their disability application is accepted should apply for both disability and retirement benefits simultaneously. Beneficiaries who weren’t aware of this option or who developed a disability after applying for retirement benefits can still apply for disability benefits after the fact. The permanent closing of some Social Security offices and temporary closing of all offices during the pandemic make it less likely that applicants will be made aware of this option or of the potential financial advantage of delaying take-up of retirement benefits.

Some economists and policymakers are concerned that disability benefits may reduce labor force participation. However, expanding government supports for workers who remain in the workforce and removing obstacles to accessing benefits can encourage employers to hire older workers by reducing employer costs associated with accommodating workers with disabilities and by helping workers who develop health conditions that interfere with work exit the workforce. Even if improving access to benefits has a net negative effect on labor force participation, disability insurance is generally welfare-improving because it helps relatively healthy people who want to work find jobs while forcing fewer people in poor health to try to keep working.

Generally speaking, expanding social insurance programs like SSDI will tend to promote the hiring of older workers. The United States is unusual among advanced economies in its reliance on employers to provide health benefits and paid leave. Older workers are more expensive to insure and more likely to take short-term disability leave due to cancer and other illnesses associated with age, though younger workers take more time off for injuries, mental illness, and family caregiving. Social insurance such as Medicare for All and paid family leave would remove the cost of health care and paid leave from employer hiring decisions as well as shield older workers from high out-of-pocket costs.

Social insurance has many potential advantages over employer-provided benefits, including spreading costs over lifetimes, pooling risk widely, reducing incentives to deny coverage or care, and taking advantage of economies of scale and government bargaining power to restrain costs. If there is not the political will for large-scale expansions of social insurance such as Medicare for All in the aftermath of the pandemic, policymakers could take more incremental approaches, such as creating a public option or Medicare buy-in, to leave less of workers’ coverage to the discretion of employers and to minimize potential disincentives to hiring older workers.

Expand Social Security to forestall a looming retirement crisis

Even in a best-case scenario, the recession will likely result in more workers retiring earlier than planned. We should expand Social Security benefits and revenue to forestall a looming retirement crisis caused by the decline of secure employer pensions in favor of do-it-yourself 401(k) plans, among other factors. It’s worth noting that Social Security and traditional pensions act as automatic stabilizers when the economy is operating below capacity due to a decline in overall spending. Social Security and traditional pensions allow workers to retire regardless of economic conditions and inject needed funds into the economy. In contrast, 401(k)-style plans tend to exacerbate business-cycle gyrations.

Introduction

An economy that’s growing on paper is not translating into better jobs for America’s workers. Despite a tight job market and strong corporate profits, paychecks have barely budged. Much of the explanation lies with eroding worker bargaining power–which is resulting in a shrinking sliver of the benefits of prosperity being shared with working families. And the deep structural racism that still pervades our economy means that workers of color and immigrants are struggling the most. Black workers face the greatest of these economic disparities, including large gaps in pay and employment. And the Trump Administration is making matters worse by rolling back the worker protection gains of recent years, scapegoating immigrants, and attacking unions and others who seek to give workers a voice on the job. More than ever, states need to lead the way in fighting back and pioneering new solutions. This agenda for America’s workers outlines model policies and best practices for enabling governors and state legislatures to protect workers in their states so that all communities can thrive and grow.

1. Jump-Start Stalled Paychecks

Raise the Stuck Minimum Wage. At the federal level and in 21 states the minimum wage has been stuck at a paltry $7.25 since 2009, causing extreme hardship for the nearly one in three U.S. workers who struggle in low-paying jobs. States representing 21% of the U.S. workforce are already gradually ramping up their minimum wages to $15 an hour—which is what single workers will soon need to afford the basics everywhere in America.

The rest of the states should do the same—and where legislatures won’t act, states should put the minimum wage on their state ballots in 2020, the way that Missouri, Arkansas, Arizona, Colorado, Washington, and Maine voters did in 2016 and 2018.

Restore Overtime Pay to Deliver a Middle-Class Raise. It used to be that if your boss asked you to put in extra hours at work, you got overtime pay in return. Not anymore. The share of salaried workers guaranteed overtime pay when they work more than 40 hours a week has plummeted from almost 63% in 1975 to less than 7% today. That’s because the salary threshold under which salaried workers are guaranteed overtime when they put in long hours hasn’t been updated in years and remains less than $24,000.

This means that millions of U.S. workers are working 50 or 60 hours a week, losing time with their families, and not getting any overtime pay for their hard work and dedication. It also means that employers aren’t hiring workers to do the extra work. The Obama administration ordered a long-overdue updating of the overtime threshold to about $48,000 a year. But a group of Republican state attorneys general blocked the increase, and the Trump Labor Department is expected to water it down.

California, New York, Pennsylvania, and Washington are already acting to deliver this long-overdue middle-class overtime raise for workers in their states. Governors and legislatures in other states should do the same–and should set a salary threshold of at least $55,234 by 2022–which is the equivalent of the 2016 Obama overtime salary threshold updated for wage growth. Importantly, in many states such as New Jersey, Michigan, Wisconsin, Minnesota, Colorado, Oregon, Montana, Pennsylvania, and Washington State, governors can expand overtime pay on their own through their state labor agencies without need for action by the legislature.

2. Get States Back in the Business of Fighting Wage Theft and Enforcing Other Worker Protections

Every week millions of workers are cheated when employers short their paychecks, force them to work off the clock, fail to pay even the minimum wage, or skirt employment laws by denying that they are employees. This type of wage theft is a national epidemic that robs U.S. workers and our economy of billions of dollars a year and hurts law-abiding employers that can’t compete with wage chiselers. But in many states, the agencies responsible for cracking down on employers that cheat their workers have been neglected and defunded. Governors and legislatures need to get their states back in the business of fighting wage theft and enforcing other worker protections, ranging from combatting independent contractor misclassification to preventing employers from defrauding the workers’ compensation system. Key best practices for restoring effective enforcement include:

• First and foremost, increasing labor agency budgets to ensure adequate staffing and enforcement capacity
• Developing strategic enforcement priorities, in partnership with worker organizations
• Cracking down on retaliation against workers who speak up
• Reviewing and updating regulations and administrative guidance—for example, to provide clear guidance on business’s responsibilities for contract workers, as detailed below.

3. Protect Contracted Workers in our Fissured Economy

A major driver of unstable and insecure work is our increasingly “fissured” economy. Today, much of the workforce for major corporations is employed indirectly through temp and staffing agencies or other contract firms, or labeled—and often mislabeled—as independent contractors. Employees wrongly treated as independent contractors are excluded from the protections of our core labor laws, such as the minimum wage, overtime, anti-discrimination protections, workers’ compensation, unemployment insurance, paid sick leave, and paid family leave. And employees working for major companies indirectly through what are often thinly capitalized, fly-by-night contractors are left wondering who’s the boss. The practice is of particular concern in sectors where workers of color are relegated to low-paid and often dangerous work, such as janitorial, delivery, home care, agriculture, landscaping, security, hospitality, trucking, transportation, and warehousing.

Corporations’ use of these work structures—subcontracting, temp and staffing, and calling workers “franchisees” or “independent contractors”—are key drivers of eroding labor standards and occupational segregation. They shift power away from workers toward corporations, and in many cases, are employed as a tactic to side-step compliance with labor laws.

As their workforces struggle with increasingly unstable work, states are responding with policy solutions to hold major companies that are the real economic actors accountable for the treatment of these workers.

Issue Clear Guidance on Business’s “Joint Employment” Responsibilities and Misclassification of Independent Contractors. Under existing employment laws, companies that use contracted workforces to staff their operations can already be held responsible as “joint employers” for those employees’ labor standards. And many purported independent contractors are, in fact, misclassified. Tightening up joint employer and independent contractor standards and improving their enforcement are key strategies for improving accountability and job standards in the fissured economy.

In 2016, the Obama Labor Department issued guidance outlining the basis for broad joint employer enforcement under the Fair Labor Standards Act, and companion guidance on independent contractor misclassification. However, the Trump Administration withdrew both. Governors and their state labor agencies should protect workers in their states against the Trump rollback by adopting similar guidance or regulations to guide agency personnel, businesses, and workers in interpreting coverage of state employment laws such as wage and hour, anti-discrimination, unemployment insurance, and workers’ compensation.

Adopt Temp and Staffing Agency Worker Protection Laws. In 2017, Illinois adopted model temporary agency worker protection legislation. It ensures that temp and staffing agencies report demographic information about the workers they hire; never charge workers for background checks, drug tests, and credit checks; notify temp workers about the types of equipment, protective clothing, and training needed to perform the job; provide transportation back from a job site if transportation was provided to the job site; and place their temporary workers into permanent positions when they become open. Other states should follow Illinois’ lead—and build on it with additional key protections to ensure that temp and staffing agency workers receive wages and benefits comparable to that received by direct employees at the companies that employ them.

Make Host Company Responsible for Labor Violations by Contractors. In 2014, California tackled the problem of fly-by-night labor contractors cheating their workers by passing new protections making host companies jointly responsible when their contractors fail to comply with minimum wage, health and safety, and workers’ compensation laws. Other states should replicate this best practice for cracking down on wage theft in our fissured economy.

4. Protect On-Demand Platform Workers and Workers Labeled–and Often Mislabeled–Independent Contractors

For decades, corporations have characterized workers as “self-employed” as a tactic for shifting economic risk downwards onto workers while maximizing revenue for investors and CEOs. In the past, it was sectors like home care, trucking, and delivery that used these tactics. Today it is also the companies that dominate the platform economy (sometimes called the “on-demand” or “gig” economy) that are aggressively seeking to shed responsibility for the employees performing work for them. By attempting to sidestep basic employment protections—from employer social insurance program contributions, to the minimum wage and overtime, to anti-discrimination and health and safety protections—these companies are leaving their workforces impoverished and vulnerable.

As the platform economy matures, the public is gaining a clearer understanding of the poor quality of these jobs. States and cities are responding by clarifying that these workers are employees covered by our nation’s baseline employment protections, and by promoting innovative sectoral solutions to improve wages and benefits for workers in sectors where work is dispatched both on- and off-platform, including transportation and domestic work. At the same time, the multi-billion-dollar platform corporations are mounting an aggressive lobbying push to try to exempt themselves from responsibility for the well-being of their workers. Governors and legislatures should adopt the following best practices to protect this growing workforce, while fighting carve-out efforts.

Clarify That the Rights and Protections of Employees Apply to Platform Workers. States should start by clarifying—through interpretation of existing laws or by amending those laws—that state employment laws (such as state minimum wage and overtime, anti-discrimination, unemployment insurance, and workers’ compensation laws) protect platform workers. This will ensure that platform workers are treated as employees, and that platform businesses are accountable for the protections that all other employers provide.

For example, in April 2018 the California Supreme Court issued a unanimous decision in the Dynamex case that will make it harder for companies, including digital platform companies, to misclassify their employees as independent contractors. Under the “ABC” tested adopted in Dynamex, businesses that seek to treat workers as independent contractors have to show that the workers are (A) free from control and direction by the hiring company; (B) perform work outside the usual course of business of the hiring entity; and (C) are independently established in that trade, occupation, or business. This test is simple, clear, and easy to enforce. More than half of the states already have the ABC test in their state unemployment insurance laws, and several including Massachusetts, New Jersey and Connecticut have adopted it for use under their wage and hour protections.

Similarly, Oregon’s labor agency issued an opinion advising that drivers at transportation network companies such as Uber and Lyft qualify as employees who are covered by the state’s employment laws. New York has ruled that platform employers are covered under the state’s unemployment insurance law. San Francisco amended its minimum wage protections to clarify that they apply to independent contractors and employees alike. The New York City Council recently passed a slate of laws meant to guarantee a $15 minimum wage to transportation network company (TNC) drivers and to stop the race to the bottom that has impoverished taxi and TNC drivers alike. Through amendment or interpretation, states should do the same for all baseline employment laws.

Promote Innovative Sectoral Solutions. In addition to clarifying that platform workers are covered by employment laws, states should promote sectoral solutions to improve jobs, including platform jobs. For example, in 2018 Seattle passed an innovative domestic worker bill of rights that requires “hiring entities” to pay domestic workers the municipal minimum wage, allow them meal and rest breaks, and protect them from retaliation. The Seattle ordinance also sets up a system for setting industry-wide standards for domestic workers, in which domestic workers themselves will have a seat at the table. Portland, Oregon has launched a similar initiative with a newly constituted board to develop standards for drivers for transportation network company (TNC) such as Uber and Lyft.

Fight Employment Protection Carve-Out Bills. On defense, states should fight back against stealth attacks on platform worker rights, such as the “marketplace platform bills” that online platform employers have passed in several states to carve their workforces out of basic protections such as the minimum wage, unemployment insurance, and workers’ compensation. The idea that it is excessive or burdensome for the multi-billion-dollar tech giants of the platform economy to provide the same basic employment protections that all other employers must follow is outrageous, and should be rejected—as Colorado and other states have done.

5. Leverage State Employment and Contracting Power to Improve Jobs and Crack Down on Mistreatment of Workers

States have a significant impact on workforce standards in their capacities as employers, and through their contracting and purchasing programs. In their own employment, states should lead by example by adopting model employment practices around fair pay and benefits. And in contracting, states should leverage their vast economic footprint to improve jobs and crack down on mistreatment of workers.

Adopt Labor Standards for Public Employees and Employees of Contractors. Governors and legislatures should start by adopting model employment practices for their own direct employees, for employees of major state-linked institutions, such as state universities, and for vendors performing state contracts. These standards should include: (1) a $15 minimum wage for state employees, state university employees, and state contractors, as states such as Massachusetts, North Carolina, and New York have done in various forms; (2) paid sick leave as the Obama Administration required for federal contractors; and (3) other core employment standards, such as “ban the box” fair hiring protections, as detailed below. In addition, preferences for contracts should be given to unionized workplaces.

Require Public Contractors to Disclose Employment and Labor Law Violations and Related Practices. In addition, states should crack down on employment and labor law violations by state contractors by adopting state responsible contracting rules, following the model of President Obama’s Fair Pay and Safe Workplaces executive order, which the Republican Congress rolled back. States can and should require companies seeking state contracts to disclose all recent federal, state, and local employment and labor law violations, ranging from wage and hour and workers’ comp, to OSHA and NLRA violations, and discourage agencies from awarding contracts to vendors with significant or repeated violations. In addition, as noted below, states should discourage the use of forced arbitration and other coercive waivers by state contractors.

6. Fight Coercive Waivers that Prevent Workers from Enforcing their Rights and Lock Them in Poverty-Wage Jobs

Bad corporate actors are increasingly using forced arbitration and other coercive waivers of worker protections to mask wrongdoing and block working people from vindicating their rights in court. According to the Economic Policy Institute, more than 60 million U.S workers are blocked from suing their employers due to forced arbitration clauses, with women, people of color, and low-wage workers disproportionately impacted. And the U.S. Supreme Court recently ruled in its Epic Systems decision that employers can force workers into private arbitration with class and collective waivers, meaning each worker has to go it alone or, more likely, not at all.

This employer-dominated process, where settlements are secret and workers are barred from joining together to seek relief as a group, allows years of abusive treatment to remain hidden. A bipartisan group of all 50 state attorneys general in 2018 called for an end to forced arbitration for sex harassment claims. But the problem extends beyond just sex harassment and equally prevents workers who have been cheated out of their paychecks from seeking justice.

Issue Forced Arbitration Executive Orders. First, governors should push back against this abusive practice by following the lead of Washington State Governor Jay Inslee and issue executive orders ensuring that tax dollars are not invested in businesses that use forced arbitration, and that companies seeking state contracts must disclose details around their use of this abusive practice.

Adopt Whistleblower Enforcement Laws. Second, to really tackle this urgent problem, governors and legislatures should restore the ability of workers and members of the public to go to court to fight wage theft, racial and sexual harassment and discrimination, and other workplace violations on behalf of the state, by adopting whistleblower enforcement laws as California has done. Model legislation, the EMPIRE Worker Protection Act, is currently pending in New York.

Ban Non-Competes and No Poaching Policies. Third, governors, legislatures, and attorneys general should fight other coercive waivers of workplace rights such as non-compete and no-poaching requirements imposed by employers on a wide swath of low-wage and other workers. These two increasingly common practices have come under growing criticism as unfair and unnecessary limits on employees’ job mobility that are contributing to stagnant wages across our economy. There is growing recognition that non-compete and no-poaching requirements are pervasive, abusive, and not necessary, as legitimate employer concerns about trade secrets held by higher-paid employees can more appropriately be addressed through non-disclosure requirements. Governors and legislatures should join California, Oklahoma, and North Dakota by prohibiting non-compete and no-poaching requirements and include a private right of action to facilitate the enforcement of such prohibitions.

7. Rebalance our Economy by Restoring Workers’ Bargaining Power

Union popularity is the highest it has been since the Great Recession: recent polls show that 62 percent of adults support unions, including an overwhelming 75 percent of young adults aged 18 to 29. With inequality at record levels and weak wage growth for most workers despite a tight labor market, there’s a growing consensus we need strong action to restore workers’ bargaining power. By joining together through unions, U.S. workers built the middle class and expanded access to good jobs, especially for women and workers of color. Strong unions raise pay and improve workplace standards not just for their members, but across the economy.

But decades of corporate-funded attacks culminated in the Supreme Court’s anti-union Janus decision in 2018, and in a spate of attacks on unions and workers by the Trump Administration and by legislatures in several states. Workers need governors and legislatures to defend the right to organize and roll back past legislative attacks.

Restore Bargaining Power for Public Sector Workers. States directly regulate collective bargaining for public sector workers, which means governors and legislatures have a significant role to play. They should step in to defend the public sector workers who deliver the vital services that sustain our communities but who are under siege and today struggle to afford the basics for their families. Governors and legislatures should use their executive and legislative powers to reverse past state attacks that restrict public sector workers’ ability to bargain collectively for fair pay, benefits, and treatment on the job, and adopt best practices to promote workable collective bargaining in the face of the Janus decision.

Support Union Efforts to Promote Good Jobs for Private Sector Workers. While states do not regulate collective bargaining for private sector workers, they can and should play a role in supporting efforts by unions in the private sector to promote good jobs. For example, in state financed and regulated sectors, such as airports, health care, and subsidized caregiver jobs, which have been characterized by low pay and poor working conditions, unions in many states are pushing to improve wages and benefits. Governors should support these efforts, as leaders in many states are doing.

Repeal Right to Work. So-called “right-to-work” laws, passed in 27 states, make it harder for workers to form strong labor unions through which they can organize and speak with one voice on the job. These laws have led to declining union membership as well as declining wages and benefits for union and nonunion workers alike. At a time of extreme inequality in our country, governors and legislators should make it easier, not harder, for workers to unite. Governors and legislatures should repeal state right to work laws currently on the books and should fight any new right to work efforts–including efforts to adopt right-to-work at the local level.

Expand Collective Bargaining Rights for Agricultural Workers and Other Workers Not Covered by the NLRA. The agricultural workers who grow the food that sustains our communities are some of the lowest paid workers in our economy. Largely migrant immigrant workers, they face not only high poverty, but grueling and dangerous working conditions. And they are excluded from the National Labor Relations Act—meaning that they have no right to join together in a union and negotiate for fair treatment. States, however, are able to regulate collective bargaining for agricultural workers, as well as other workers such as domestic workers and independent contractors who are excluded from federal protections. States should follow the lead of California, which more than 40 years ago adopted the California Agricultural Relations Act. A campaign for similar agricultural workers labor relations legislation is underway in New York.

8. Promote Fair Hiring for People with Arrest or Conviction Records

Roughly one in three adults in the U.S. has an arrest or conviction record that can show up on a routine criminal background check for employment, undermining the job prospects of the 70 million men and women who have been caught up in the criminal justice system. This legacy of mass incarceration has an especially devastating impact on the employment prospects of people of color, who are 40 percent less likely than white applicants with a record to receive a positive response from a prospective employer.

Adopt Fair Chance Hiring. In response, 33 states, including Georgia, Kentucky, Louisiana, North Carolina, Tennessee, and Virginia, have adopted “ban the box” policies to open up job opportunities in state and local government for people with arrest or conviction records and set an example for private sector employers. Eleven states, including large states like California, and more than a dozen major cities across the U.S. extend fair chance hiring protections to private sector employers, in addition to the public sector. Governors from states that haven’t yet joined them should start by issuing executive orders adopting this reform for all state hiring. Then they and their legislatures should push for legislation to extend this best practice to the private sector, as more and more states and cities are doing.

Remove Occupational Licensing Barriers for People with Records. Today, more than 25 percent of U.S. workers must obtain a state license or certification before they can work in their chosen occupation, and onerous criminal background check restrictions often accompany these state licensing mandates. With broad bipartisan support, over the past year about a dozen states have taken bold steps to remove unfair restrictions against hiring people with records from their occupational licensing laws. In 2018, the governors of several states (Michigan, New Mexico, and Pennsylvania) took executive action directing licensing boards or other state entities to take action to address unnecessary restrictions that limit qualified people from fairly competing for jobs in their chosen professions. Governors and legislatures should follow their lead with executive action and legislation to remove unnecessary licensing obstacles to employment for people with records.

Adopt a Clean Slate Policy for People with Records. Having even a minor criminal record can be a life sentence to poverty; in addition to being a barrier to employment, it is increasingly a barrier to housing and even education as landlords and colleges use background checks to screen applicants’ criminal records. While most states allow people to petition to have their records expunged or sealed, only a tiny fraction of people eligible ever get the relief they need because they can’t afford a lawyer, pay the court fees, or figure out how to navigate the complex court petition process. Many are not even aware it’s an option. Governors and legislatures should adopt “clean slate” policies that provide for automatic record clearing once someone remains crime-free for a designated period of time. People with criminal records who have remained crime-free for four to seven years are no more likely than the general population to commit a new crime. Pennsylvania enacted clean slate legislation in 2018, and several other states, including Colorado, Michigan, and South Carolina, are considering it. A majority of voters–across party, racial, gender, and education lines–support the policy.

9. Fight Racial and Gender Discrmination on the Job and Combat High Unemployment in Communities of Color

Structural racism and gender discrimination still pervade our economy and labor markets. African American men are paid just 71 cents for each dollar that white men are paid. African American women are paid even less—just 66 cents, while white women earn 78 cents. And despite a tight labor market and record low unemployment in general, the African American unemployment rate remains nearly double the white rate.

The drivers of these deep racial and gender disparities in our job market include discrimination in hiring and pay, occupational segregation, and weak enforcement systems. Governors and legislatures should work to dismantle them with a multi-pronged approach.

Banning Employers from Asking About Salary History. There is growing national recognition that the common practice of employers basing employees’ pay in part on their salary history perpetuates unequal pay for women and workers of color, since gender and racial pay gaps are often present, even among new entrants to the workplace. In response eleven states and nine cities have banned employers from inquiring about salary history and basing compensation on it. Governors and legislatures should follow their example and adopt these sensible prohibitions–and require that employers seeking state contracts refrain from such practices.

Strengthening Civil Rights Enforcement. State human rights agencies are a first line of defense in the fight against discrimination on the job. But many have seen their budgets and staff slashed or stagnant and as a result have long backlogs and little capacity to engage in strategic enforcement. At the same time, human rights agencies are reporting a spike in discrimination and harassment complaints—a trend that is likely fueled by President Trump’s hostility to immigrants and people of color. Governors and legislatures should rebuild their state human rights agencies by restoring adequate staffing and budgets—and promoting strategic, targeted enforcement programs by these agencies to help them more effectively combat discrimination in hiring, promotions and pay, as well as workplace harassment. This includes implementing reporting systems that create greater pay transparency and deepening partnerships with community based organizations. In order to bolster enforcement of civil rights laws, states should also ensure that localities are not preempted from expanding protections beyond what the state law may provide and enforcing local anti-discrimination laws consistent with the state protections.

In addition, in recent years states have begun to enact stronger laws prohibiting unequal pay for similar work—but have not always included race-based pay inequality in these protections. California recently expanded its equal pay law, the California Fair Pay Act, to include race-based pay inequality. Governors and legislatures in other states should follow California’s lead and adopt state-of-the-art equal pay protections—and be sure that they tackle pay inequality based not just on gender, but on race and other protected statuses as well.

Promoting Targeted Hiring of Workers from High Unemployment Communities. Many cities are successfully using “targeted hiring” or “first source hiring” programs to ensure that tax-payer funded projects expand access to jobs for workers from communities with high unemployment rates. Given the disproportionately higher rates of unemployment in Black communities that have been shut off from economic opportunity for decades, these targeted hiring programs are an important step toward greater employment equity. For example, on publicly funded development projects, these programs typically require that construction contractors or end-user occupants like stadiums, hotels or retailers on development projects target a share of their hiring at workers from high-unemployment or high-poverty zip codes, by partnering with community-based job training and referral agencies. Other cities are encouraging publicly linked “anchor institutions” like universities or hospitals to enter into similar partnerships to fill health sector and service jobs, or requiring that major government service contractors hire from targeted, high unemployment communities. Examples include New York City’s multi-pronged HireNYC Program, which combines targeted hiring for construction jobs, end-user jobs on subsidized development projects, and jobs with major city service contractors.

Governors and legislatures should scale these programs up to the state level by adopting targeted hiring programs for state-linked projects and institutions that generate significant numbers of jobs, such as state-financed economic development, state service contracting, direct state civil service hiring, and state-financed institutions like universities and hospitals. In all of these spheres, states should develop systems for filling a portion of jobs with workers from high-unemployment zip codes, and develop a system of partnerships with community-based job training and referral agencies to make it work.

Inclusive Procurement and Contracting. Another key strategy for ensuring that employment and wealth-building opportunities generated by states’ infrastructure and contracting programs are shared equitably with disadvantaged communities is inclusive procurement and contracting. Inclusive procurement and contracting involves a suite of policies to ensure that minority and women-owned businesses enterprises (M/WBE’s) are fairly represented among firms selected for contracts or subcontracts on state-financed procurement and infrastructure projects. Because minority-owned firms are more likely to employ a diverse workforce, inclusive contracting programs can be an effective approach for ensuring that workers of color benefit from the jobs generated by state spending. Governors and legislatures should review their contracting programs and implement recommended strategies for ensuring equitable inclusion of M/WBE’s. At the same time, they should ensure that state-financed contracting and infrastructure projects are covered by strong labor standards, including prevailing wages, community workforce agreements for construction projects, and $15 minimum wages for service contracting, to ensure that they generate quality jobs and do not undercut high road standards.

10. Protect Workers’ Health and Safety

Nearly 50 years after Congress adopted the Occupational Safety and Health Act (OSHA) requiring employers to provide safe workplaces, more than 5,000 U.S. workers are killed on the job every year, and nearly three million are seriously injured. Many low-wage jobs are dangerous jobs, including jobs in the poultry and meat industries, agriculture, construction, and home care, where workers suffer much higher rates of serious job injuries. Yet the Trump Administration is rolling back workplace health and safety protections, leaving workers even more vulnerable.

Adopt Responsible State Health and Safety Contracting. Governors and state legislatures should fight these rollbacks by promoting model protections for workers. For example, Massachusetts is considering a model responsible contracting law for health and safety. It requires contractors and subcontractors bidding on state-funded projects to submit their health and safety violations histories—and bars contracting with companies with poor records. Legislatures and governors using their executive authority over contracting should adopt this model.

Stronger State Workplace Protections on Heat Exposure. With climate change, heat exposure is emerging as a very serious workplace health hazard in sectors from agriculture to day labor. But currently there are few standards or protections. Governors and legislatures should adopt new standards and programs to provide stronger protections for workers exposed to dangerous levels of heat, especially farm workers but also workers in construction, manufacturing, and warehousing—all sectors where workers of color and immigrants are concentrated.

Strengthen Workers’ Compensation Laws. Over the past two decades, state legislatures have engaged in a race to the bottom by hollowing out their workers’ compensation laws, resulting in unfair, weak, or nonexistent benefits for injured workers. Governors and legislatures should work together to prevent any further weakening of benefits and coverage–especially since workers’ compensation premiums and benefits are now at a 30 year low. Key workers’ compensation reforms that are needed in most states include: (1) strong anti-retaliation protections for injured workers; (2) insurance coverage for prompt medical care in contested cases; (3) extending coverage to all workers, including domestic workers, farm workers, and temporary workers; and (4) ensuring that workers have the right to choose their own doctor.

Fighting Sexual Violence on the Job. Low-wage workers such as janitors, hotel room cleaners, and waitresses are especially vulnerable to sexual assault and violence in the workplace, as many work in isolation and feel powerless to speak up. California responded in 2016 with the Property Services Worker Protection Act, which helps protect janitors from rape on the job by mandating sexual harassment and assault prevention training, and registration of property services contractors. Similarly, cities including Chicago, Seattle, Oakland, and Long Beach, California have enacted measures to protect hotel room cleaners that require that they be provided panic buttons to use when they enter rooms alone, and also establish workload standards to protect room cleaners. Governors and legislatures should follow the lead of these states and cities in protecting low-wage workers from assault at work.

11. Promote Quality Caregiver Jobs

Our nation is facing a care crisis. States are struggling to provide adequate access to quality affordable home care and child care services for seniors, people with disabilities, and working families, at the same time that they struggle to improve the quality of these vital caregiving jobs. A big part of the problem is chronic underinvestment in the home care and child care workforces, which are some of the lowest paid in our economy, promoting high poverty rates and workforce instability among these vital caregivers–many of them women of color and immigrants.

Governors and state legislatures should join the states and cities that are tackling this crisis with new approaches that expand access to these vital services for all families, while investing in upgrading the jobs.

Expand, Enforce and Adequately Fund Minimum Wage, Overtime and Paid Sick Days Protections for Caregivers. States finance and regulate a large swath of the home care workforce through their Medicaid long-term services and supports programs. Governors and legislatures should ensure that these vital caregivers receive an adequate minimum wage, overtime pay coverage, and paid sick days protections, along with any other benefits. First, they should ensure that their Medicaid home care programs are implementing–and adequately budgeting for–the Obama Labor Department’s 2015 “companionship” rule that finally extended federal minimum wage and overtime protections to homecare workers. For example, they should follow the lead of states such as California, which budgeted extra funding for its consumer-directed Medicaid funded home care program to account for overtime and travel time between consumers. States must also ensure that consumers have access to quality and adequate home care services that allow for them to remain at home and within the community. Second, they should guarantee a $15 minimum wage for Medicaid home care workers–which governors can do by executive action and negotiate funding for as part of the budget process, as Massachusetts Governor Charlie Baker did. Third, they should extend similar $15 minimum wage to workers in the state’s subsidized family child care provider program–as Massachusetts also recently did as part of a 2018 minimum wage package. Lastly, states should ensure that their state laws cover home care, child care, and other domestic workers–and that they provide robust enforcement mechanisms to ensure workers’ rights are upheld.

Raise Caregiver Wages by Increasing Medicaid Reimbursement Rates for Long-term Care Services. The rates at which Medicaid reimburses providers for the long-term care services they provide seniors and people with disabilities can factor into caregivers’ low-wages. In many states, Medicaid reimburses homecare expenses at hourly rates that may provide too little to adequately pay these workers living wages or even cover homecare agency operating expenses. Increasing these reimbursement rates to ensure workers can earn living wages and have benefits, and requiring that the majority of the increase go to workers’ wages, will ensure that homecare jobs can provide workers with a decent living. States should also ensure that their consumer-directed rates are adequate to ensure home care workers make a living wage.

Create Universal Home Care with Quality Jobs. In Maine, an innovative campaign seeks to establish a Universal In-Home Care Program. It would expand access to subsidized home care services for all Mainers, regardless of income, while simultaneously creating a program board charged with improving wages, benefits, and working conditions for the home care workforce. The campaign to adopt this policy by ballot initiative fell short in 2018, but advocates are hopeful about pursuing similar legislation in the state legislature in 2019.

Expand Affordable Child Care with Quality Jobs. In Alameda County, California, a similar campaign is proposing to create an expanded subsidized child care program for lower income and middle-class families. It combines opening up affordable child care to more of the working class, with upgrading subsidized child care jobs significantly, financed by a half-cent sales tax increase.

12. Update and Defend Social Insurance Programs for All Workers

Vital social insurance programs on which working families rely have been under attack by the Trump Administration and by governors and legislatures in many states in recent years. As new governors and legislatures take office, they should take strategic action to restore these vital programs for workers in their states–and prepare their economies for the next recession or natural disaster.

Repeal Medicaid Work Requirements. Several states have instituted or are seeking federal permission to impose short-sighted and punitive work requirements on residents seeking health care coverage under Medicaid. These policies aim to take basic health care away from low-income workers and people with disabilities, many of whom work but are unable to keep up with the burdensome documentation requirements under such programs. New governors and legislatures should reverse such waivers and restore health care access for low-incomes residents in their states.

Restore a Strong Unemployment Insurance System. After poor financing decisions caused more than 40 states to bankrupt their unemployment insurance (UI) trust funds during the Great Recession, many states slashed their UI programs. Even before the recession, the UI program needed reform; that’s even truer today, with only about one in four unemployed workers receiving UI benefits. With another cyclical recession likely in the coming years, states should act to restore hard-earned benefits under their programs to protect workers, their families, and their communities from the inevitable downturn. NELP’s extensive toolkit details a full range of key UI reforms that many states have already implemented. Key reforms include a minimum uniform 26 weeks of benefits, work-sharing programs to preserve jobs in times when companies are struggling or during a recession, and reforms to provide UI protection for part-time workers who lose their jobs–a significant segment of the workforce, where women and low-wage workers are concentrated.

Prepare for Disaster Unemployment Assistance. For states that face regular natural disasters, modernizing Disaster Unemployment Assistance (DUA) systems is crucial for helping workers who lose their jobs during disaster recover from such devastating losses. All states affected by disasters should make key reforms. First, they should suspend or get rid of the “waiting week,” which only delays support to workers. Second, given the difficult realities and limitations facing those seeking work in a disaster area, the affected states and territories should significantly relax or suspend their work-search mandates. States should also take steps to boost UI benefits and authorize “non-charge” benefits paid as a result of a disaster.

Protect Public Sector Pensions. As the United States faces a looming retirement crisis, many states are looking to slash guaranteed retirement plans for the one segment of the population that still has a modest but secure benefit–public sector workers. While some states skipped contributions to their pension funds, workers have always paid their share. But some policymakers are asking workers to sacrifice their future for the mistakes past lawmakers made. A solid retirement plan is a public good. It helps states and cities attract and retain great teachers, firefighters, and caregivers while also providing a boon to the local economy when those workers retire in their communities. States should maintain and fund secure public sector retirement plans–and work simultaneously to develop new retirement options for private sector workers.

13. Empower Local Communities by Fighting Preemption

In recent years, big corporations and their lobbyists have not only blocked state efforts to raise the minimum wage, guarantee paid sick days, or address other worker needs, but when cities and counties have tried to tackle these problems, the corporations and their lobbyists have stepped in to tie their hands with “preemption laws.” Like voter disenfranchisement and political gerrymandering, these preemption laws are part of a corporate strategy by groups like the Koch Brothers-funded American Legislative Exchange Council (ALEC) to stymie progressive action. These preemption efforts are disproportionately impacting efforts by communities of color to improve local economic conditions and address specific localized concerns. Governors and legislatures should pledge to block or veto any new attempts to limit local power to address worker needs, and should work to roll back existing limits. For example, they should follow the lead of Colorado Governor-elect Jared Polis, who is backing a campaign in the legislature to repeal wage preemption in that state. At the same time, however, state leaders should fight any local efforts to undercut established worker protections, such as illegal local “right to work” efforts, or punitive local anti-immigrant measures.

14. Enable Workers to Balance Work and Family Demands by Guaranteeing Paid Sick Days, Paid Family Leave and Fair Scheduling

The U.S. lags far behind the rest of the world in ensuring basic protections to allow workers to balance the demands of work and family life such as earned paid sick days and paid family leave. And the growing problem of irregular and unpredictable work schedules, fueled by new scheduling technology, is posing serious hardship for working families. Governors and legislatures follow the lead of the growing numbers of states and cities that are responding by guaranteeing paid sick days and paid family leave, and by adopting fair workweek scheduling legislation.

Paid Sick Days. More than 34 million workers in this country don’t have a single paid sick day — and each time they take needed time off, they risk their families’ economic security and jeopardize the public’s health. Governors and legislatures should join the eleven states and dozens of cities that have adopted earned paid sick days laws to ensure that all workers have access to this most basic of protections.

Paid Family Leave. Sixty percent of the U.S. workforce does not have access to paid family leave when they need to take time off from work after the birth or adoption of a baby or in case of illness. As a result, millions of workers either cannot take the time they need, or must take it with no pay–and with no guarantee that their employer will hold their job for them when they return. In response, more and more states are establishing paid family leave insurance programs to provide workers with a share of their wages when they need time to care for a family member with a serious health condition, bond with a new child or deal with their own serious medical issue. Governors and legislatures should follow their lead and adopt paid family leave for their states.

Fair Workweek. There is growing recognition that unpredictable, unstable, and often insufficient work hours are a key problem facing many U.S. workers, particularly those in low-wage industries. Volatile hours not only mean volatile incomes, but add to the strain working families face as they try to plan ahead for child care or juggle schedules in order to take classes, hold down a second job, or pursue other career opportunities. The problem has been made worse by new scheduling technology that has enabled many retail and fast food employers to adopt last minute, “just in time” scheduling.

In response, states and cities are starting to adopt “fair workweek” laws that provide workers with greater stability, predictability, and flexibility in their work schedules. In many cases, they also require employers to give part-time staff opportunities to increase their hours before adding new staff. Governors and legislatures should follow their lead and push for fair workweek legislation to protection workers in their states.

15. Protect Immigrant Workers

Immigrant workers are vital members of our communities and contributors to state economies. But President Trump is bent on scapegoating immigrants and driving them back into the shadows. Governors and legislatures should take a strong stand to support immigrant workers rights and protect workers against immigration-based retaliation.

Ensure That Employment and Labor Laws Protect All Workers and Fight Retaliation. States should affirm that their employment and labor laws, such as the minimum wage and workers compensation, protect all workers regardless of immigration status. All workers need basic protections–and failing to enforce the law for undocumented workers hurts responsible employers and documented workers alike. It also creates incentives for employers to exploit undocumented workers in their workforces. States should also adopt strong anti-retaliation policies to protect workers who come forward against employer retaliation or ICE interference.

Provide Guidance for Employers About Immigration Status Verification and Worksite Immigration Enforcement. Employers need guidance on immigration status issues to help them avoid over-reacting and inadvertently cooperating with anti-immigrant attacks. States should provide guidance on best practices for responding to ICE workplace enforcement and immigration status verification, including educating employers that there is no need to re-verify DACA or TPS holders’ workplace authorization, and that Social Security Administration no-match letters are not proof of undocumented status.

Protect Civil Rights and Public Safety. States should also promote civil rights and public safety in their communities by keeping immigration enforcement out of policing. As law enforcement leaders have explained, allowing police departments and other state or local officials to cooperate with ICE seriously compromises public safety by eroding immigrant communities’ trust in law enforcement. Governors and legislatures should prevent state taxpayer money from being used to enforce the Trump Administration’s xenophobic agenda and refuse to subsidize immigration enforcement or incarcerate individuals because of an ICE detainer.

Promote Access to Drivers’ Licenses. States should also expand access to drivers’ licenses for all workers, regardless of immigration status to promote public safety. Ensuring that all workers can obtain drivers’ licenses improves traffic safety, reduces uninsured motorist accidents, and benefits the economy.

16. Protect Savers from being Ripped Off by Wall Street

In 2016, the Obama Labor Department issued a new rule requiring financial professionals to put their customers’ interests first when providing retirement investment advice. This “fiduciary rule” required financial professionals to rein in conflicts of interest, like kick-backs and sales contests, that encouraged them to cheat their clients. Every year, retirement savers alone lose more than $17 billion due to financial advisors’ conflicts of interest, and the losses are much greater when one considers all accounts and all products.

Wall Street and its allies in the insurance industry mounted a relentless attack on this common-sense ban on predatory investment practices, since it would have cut into their profits. They challenged the rule in court and found a sympathetic panel that was willing to block the rule. Siding with the industry, the Trump Administration abandoned the rule. Now the Securities and Exchange Commission is in the process of replacing it with a far weaker, watered-down rule backed by the same industry groups that fought against real protections in the first place.

Governors, legislatures, and state regulators can and should step in to protect investors against this Trump rollback by adopting a strong fiduciary rule for financial professionals in their state. Unlike the Labor Department rule, which applied only to retirement accounts, and the SEC proposal, which would apply only to securities accounts, states can adopt a fiduciary rule that applies to all financial professionals who provide investment advice, including both securities and insurance professionals, and that protects all investors, retirement and non-retirement alike.

17. Promote Economic Development That Generates Broadly Shared Prosperity

After decades of income inequality and wage stagnation, it is increasingly clear that topline economic growth does not produce bottom line prosperity. As a result, states need to pursue equitable economic development policies that spread the benefits of GDP growth more widely, especially to communities of color, which thanks to historically-rooted patterns of discrimination continue to lag the national average in wages, wealth, and social mobility. This kind of equitable economic development is focused not just on topline GDP growth or the total number of jobs in a community, but on the quality of the jobs—their wages, benefits, and prospects for upward mobility. It focuses on boosting the capacity of communities to build their own economic momentum and seeks to bring marginalized and disconnected workers into the labor market, rather than relying solely on outside capital to fuel growth and prosperity. In turn, this requires leveraging existing assets that train these workers, boost the productivity and innovative capacity of existing businesses, and leverage the critical supply chains and industry clusters that globally competitive businesses need to thrive. And it means intentionally extending these efforts into underserved communities, including rural areas and communities of color. This kind of economic development starts with the basic objective of making sure these programs actually deliver on their promises through better accountability and transparency, and then takes a step forward by incorporating equitable development practices into the state economic policy toolbox.

Improve Transparency for Economic Development. Each year, states spend tens of billions of dollars in economic development subsidies designed to lure businesses—and theoretically jobs—to their states. As the highly publicized Amazon HQ2 tax-break sweepstakes has revealed, big economic development decisions are made largely behind closed doors, leaving taxpayers with little to no input. When taxpayers do seek to participate, they often find that they cannot obtain even the most basic information about the deals being considered. The result is that a company’s announcement to build in an area, accompanied by the government’s announcement of a large subsidy package, is often the first official word the public hears about a development project.

Far too frequently, press announcements of economic development deals tout the projected number of jobs that will be created, but the fine print is full of loopholes. Even if a development deal is tied to jobs created or dollars invested, other big accountability safeguards are often missing. Will the actual jobs created and wages paid be publicly disclosed? Will the jobs pay living wages? Provide health and retirement benefits? Give local residents a chance to get hired? Be accessible via public transit? Be environmentally responsible? Create affordable housing? Preserve open space?

All too often, governments overspend on development deals, starving the public services communities rely on. Revenue lost to tax breaks would otherwise improve schools, roads, transit, public safety, and other public services that benefit all employers and working families.

Require Every Economic Development Project to Actually Create the Promised Jobs Before Awarding Public Dollars. All too often, companies never deliver on the jobs and investment they’ve promised in exchange for public subsidies. Governors and legislatures should protect taxpayers and ensure these programs are effective by requiring every company that receives a subsidy award to actually create the jobs they promised before they receive any public dollars. In some states, these programs also “clawback” or take back any public dollars that were given to companies once the company fails to meet job creation milestones.

Disclose Economic Development Deals. In many states, governors will inherit an incentive system where economic development deals are poorly disclosed. According to Good Jobs First, only one in four major state development programs reports on the number of jobs actually created or workers trained, and only one in eleven reports on wages actually paid. Alabama, Georgia, Hawaii, Idaho, Kansas, Maine, Nebraska, New Hampshire, New Mexico, Nevada, South Carolina, and South Dakota all rank in the bottom 15 among the states according to a national survey by the group. Governors should use their administrative powers to disclose the costs and benefits of every economic development deal, online, as has been common practice in many states for many years. Since economic development is an executive branch function, many states have first disclosed incentive deals pursuant to executive action. Some have later codified the practice in law, but legislation is not typically necessary.

Adopt a Unified Economic Development Budget (UEDB). UEDBs compile every kind of state spending for economic development: tax expenditures, program and agency appropriations, grants, loans, and even workforce development. Invariably, they show that tax breaks are literally the bottom of the iceberg, bigger than appropriations by ratios of 4:1, 6:1 and even higher. Yet tax breaks are far less likely to be disclosed, receive a performance audit, or be sunsetted. Governors should require the state Revenue Department to publish a Unified Economic Development Budget, so that lawmakers can see the big picture and make sure priorities are correctly balanced.

Report Tax Revenue Lost to Corporate Tax Breaks. In 2015, the Governmental Accounting Standards Board (GASB) issued Statement 77 on Tax Abatement Disclosures, which requires most localities (including school districts) and states to disclose the amount of tax revenue they lose annually to economic development tax abatement programs. Only one year’s data is out yet, however compliance with Statement 77 is uneven so far. Governors should propose legislation, or when possible direct the state auditor, comptroller, or treasurer to improve compliance with Statement 77, and also to put the new disclosures online. Bernalillo County, New Mexico, issued a recent report than can serve as a model for cities, counties and states around the country.

Combine Equitable Development Practices with Traditional Economic Development Tools Like Business Incentives. Alongside strong accountability measures, traditional tools like business incentives can be dramatically improved by adopting key equitable development practices. The first is to target incentives to retaining and supporting existing, locally-owned businesses, rather than relying on branch plants and headquarters from outside the state. Dollars spent on locally-owned businesses have long been understood to have a bigger economic impact as they circulate through the local economy than mobile firms based in other states.

A second key strategy is to direct incentives to firms in key supply chains or in industries supported by comprehensive development strategies that combine workforce training, research and development, and small business development to build out entire sectors. The growth of biotech and aerospace in North Carolina are good examples. These sector development efforts have been shown to improve the job creation and investment outcomes for the incentive projects in those industries.

Connect Marginalized Workers to Jobs Through First Source Hiring Agreements with Employers. As discussed above, targeted or first source hiring is a policy designed to ensure that private sector businesses and local governments recruit and hire disadvantaged employees. Local governments and community-based organizations across the country have successfully used first source hiring as a way of targeting training, economic development, public jobs, and public contracts to residents of the economically distressed areas or neighborhoods that most need job training and employment opportunities. First source hiring policies often require businesses receiving economic development incentives to give preference in their pre-employment screening and hiring processes to potential employees recommended by a specific, named intermediary, often a community college or community-based training organization. For example, a local community college could recruit students from a distressed neighborhood into a customized training program created for a specific employer. The students receive customized training and then are given preference for interviews and screening with the employer once they’ve completed the program.

18. Promote and Fund Affordable Housing

There is an unprecedented national housing affordability, habitability, and eviction crisis that requires a bold statewide response. Across the country, nearly 11 million renters pay over half of their income for rent and utilities, and tenants, manufactured homes residents, and low-income homeowners are rising up for housing justice in response. The affordable housing and eviction crises have grown significantly worse with no minimum wage worker able to afford a 2 bedroom apartment anywhere in the country without paying more than 30% of their income in rent. At the same time, the vast majority of tenants are without rent or eviction protections, leaving them at the whim of corporate landlords to raise rents, decreasing families’ economic security. According to the Eviction Lab, one out of four severely rent-burdened families spends over 70% of their income on rent and utilities and only one in four families who qualify for affordable housing receive housing assistance.

Key best practices for state based interventions to significantly address the affordable housing crisis and alleviate evictions are (1) expanding state funding for affordable housing, and (2) passing a comprehensive program of tenant protections including rent controls and just cause eviction protections.

Fund Affordable Housing. The greatest housing crisis exists for the lowest income renters who are most at risk of homelessness and displacement. States can alleviate the housing crisis through fully funding public housing and preserving at-risk affordable housing, as well as through fully funding housing trust funds, rental assistance programs, and state-based affordable housing programs. States should prioritize housing programs for both preservation and new construction for low-income and extremely low-income tenants, defined as those earning under 50% of Area Median Income, with a priority for those earning under 30% of Area Median Income.

Protect Tenants. The vast majority of renters live in unregulated housing without tenant protections, leaving them at great risk to housing insecurity. States should expand and strengthen tenant protections through implementing state-based rent control and just cause eviction protections to limit rents and increase eviction protections—and by repealing preemption to empower cities to adopt such protections. Rent control is a system of tenant and rent protections that can curb the crisis of evictions, displacement, and housing unaffordability. It protects tenants from profiteering landlords through a combination of rent restrictions, just cause eviction protections, and a system of holding landlords accountable for poor and uninhabitable conditions. Beyond rent control, states should increase resources to legal services for eviction defense; implement renters’ tax credits; create anti-discrimination protections for those using Section 8 and other assistance programs; and fund robust code enforcement to improve conditions and services for substandard and uninhabitable housing.

Paul Sonn, National Employment Law Project (NELP), psonn@nelp.org

Naomi Walker, Economic Analysis and Research Network (EARN),
Economic Policy Institute (EPI), nwalker@epi.org

A broader understanding of the importance of student behaviors and school climate as drivers of academic performance and the wider acceptance that schools have a role in nurturing the “whole child” have increased attention to indicators that go beyond traditional metrics focused on proficiency in math and reading. The 2015 passage of the Every Student Succeeds Act (ESSA), which requires states to report a nontraditional measure of student progress, has codified this understanding.

The vast majority of U.S. states have chosen to comply with ESSA by using measures associated with student absenteeism—and particularly, chronic absenteeism. This report uses data on student absenteeism to answer several questions: How much school are students missing? Which groups of students are most likely to miss school? Have these patterns changed over time? And how much does missing school affect performance?

Data from the National Assessment of Educational Progress (NAEP) in 2015 show that about one in five students missed three days of school or more in the month before they took the NAEP mathematics assessment. Students who were diagnosed with a disability, students who were eligible for free lunch, Hispanic English language learners, and Native American students were the most likely to have missed school, while Asian students were rarely absent. On average, data show children in 2015 missing fewer days than children in 2003.

Our analysis also confirms prior research that missing school hurts academic performance: Among eighth-graders, those who missed school three or more days in the month before being tested scored between 0.3 and 0.6 standard deviations lower (depending on the number of days missed) on the 2015 NAEP mathematics test than those who did not miss any school days.

Introduction and key findings

Education research has long suggested that broader indicators of student behavior, student engagement, school climate, and student well-being are associated with academic performance, educational attainment, and with the risk of dropping out.1

One such indicator—which has recently been getting a lot of attention in the wake of the passage of the Every Student Succeeds Act (ESSA) in 2015—is student absenteeism. Absenteeism—including chronic absenteeism—is emerging as states’ most popular metric to meet ESSA’s requirement to report a “nontraditional”2 measure of student progress (a metric of “school quality or student success”).3

Surprisingly, even though it is widely understood that absenteeism has a substantial impact on performance—and even though absenteeism has become a highly popular metric under ESSA—there is little guidance for how schools, districts, and states should use data about absenteeism. Few empirical sources allow researchers to describe the incidence, trends over time, and other characteristics of absenteeism that would be helpful to policymakers and educators. In particular, there is a lack of available evidence that allows researchers to examine absenteeism at an aggregate national level, or that offers a comparison across states and over time. And although most states were already gathering aggregate information on attendance (i.e., average attendance rate at the school or district level) prior to ESSA, few were looking closely into student-level attendance metrics, such as the number of days each student misses or if a student is chronically absent, and how they mattered. These limitations reduce policymakers’ ability to design interventions that might improve students’ performance on nontraditional indicators, and in turn, boost the positive influence of those indicators (or reduce their negative influence) on educational progress.

In this report, we aim to fill some of the gaps in the analysis of data surrounding absenteeism. We first summarize existing evidence on who misses school and how absenteeism matters for performance. We then analyze the National Assessment of Educational Progress (NAEP) data from 2003 (the first assessment with information available for every state) and 2015 (the most recent available microdata). As part of the NAEP assessment, fourth- and eighth-graders were asked about their attendance during the month prior to taking the NAEP mathematics test. (The NAEP assessment may be administered anytime between the last week of January and the end of the first week of March, so “last month” could mean any one-month period between the first week of January and the first week of March.) Students could report that they missed no days, 1–2 days, 3–4 days, 5–10 days, or more than 10 days.

We use this information to describe how much school children are missing, on average; which groups of children miss school most often; and whether there have been any changes in these patterns between 2003 and 2015. We provide national-level estimates of the influence of missing school on performance for all students, as well as for specific groups of students (broken out by gender, race/ethnicity and language status, poverty/income status, and disability status), to detect whether absenteeism is more problematic for any of these groups. We also present evidence that higher levels of absenteeism are associated with lower levels of student performance. We focus on the characteristics and outcomes of students who missed three days of school or more in the previous month (the aggregate of those missing 3–4, 5–10, and more than 10 school days), which is our proxy for chronic absenteeism.4 We also discuss data associated with children who had perfect attendance the previous month and those who missed more than 10 days of school (our proxy for extreme chronic absenteeism).

Given that the majority of states (36 states and the District of Columbia) are using “chronic absenteeism” as a metric in their ESSA accountability plans, understanding the drivers and characteristics of absenteeism and, thus, the policy and practice implications, is more important than ever (Education Week 2017). Indeed, if absenteeism is to become a useful additional indicator of learning and help guide effective policy interventions, it is necessary to determine who experiences higher rates of absenteeism; why students miss school days; and how absenteeism affects student performance (after controlling for factors associated with absenteeism that also influence performance).

Major findings include:

One in five eighth-graders was chronically absent. Typically, in 2015, about one in five eighth-graders (19.2 percent) missed school three days or more in the month before the NAEP assessment and would be at risk of being chronically absent if that pattern were sustained over the school year.

• About 13 percent missed 3–4 days of school in 2015; about 5 percent missed 5–10 days of school (between a quarter and a half of the month); and a small minority, less than 2 percent, missed more than 10 days of school, or half or more of the school days that month.
• We find no significant differences in rates of absenteeism and chronic absenteeism by grade (similar shares of fourth-graders and eighth-graders were absent), and the patterns were relatively stable between 2003 and 2015.
• While, on average, there was no significant change in absenteeism levels between 2003 and 2015, there was a significant decrease over this period in the share of students missing more than 10 days of school.

Absenteeism varied substantially among the groups we analyzed. In our analysis, we look at absenteeism by gender, race/ethnicity and language status, FRPL (free or reduced-price lunch) eligibility (our proxy for poverty status),5 and IEP (individualized education program) status (our proxy for disability status).6 Some groups had much higher shares of students missing school than others.

• Absenteeism by poverty and disability status. Poor and somewhat poor students (those who qualified for free lunch or for reduced-price lunch) and students with disabilities (those who had individualized education programs, or IEPs) were much more likely than their more affluent or non-IEP peers to miss a lot of school.
  • Twenty-six percent of IEP students missed three school days or more, compared with 18.3 percent of non-IEP students.
  • Looking at poverty-status groups, 23.2 percent of students eligible for free lunch, and 17.9 percent of students eligible for reduced-price lunch, missed three school days or more, compared with 15.4 percent of students who were not FRPL-eligible (that is, eligible for neither free lunch nor reduced-price lunch).
  • Among students missing more than 10 days of school, the share of free-lunch-eligible students was more than twice as large as the share of non-FRPL-eligible students (2.3 percent vs. 1.1 percent). Similarly, the share of IEP students in this category was more than double the share of non-IEP students (3.2 percent vs. 1.5 percent).
• Absenteeism by race/ethnicity and language status. Hispanic ELLs (English language learners) and Native American students were the most likely to miss three or more days of school (24.1 and 24.0 percent, respectively, missed more than three days of school), followed by black students (23.0 percent) and Hispanic non-ELL and white students (19.1 and 18.3 percent, respectively). Only 8.8 percent of Asian non-ELL students missed more than three days of school.
  • Perfect attendance rates were slightly higher among black and Hispanic non-ELL students than among white students, although all groups lagged substantially behind Asian students in this indicator.
  • Hispanic ELL students and Asian ELL students were the most likely to have missed more than 10 school days, at 3.9 percent and 3.2 percent, respectively. These shares are significantly higher than the overall average rate of 1.7 percent and than the shares for their non-ELL counterparts (Hispanic non-ELL students, 1.6 percent; Asian non-ELL students, 0.6 percent).

Absenteeism varied by state. Some states had much higher absenteeism rates than others. Patterns within states remained fairly consistent over time.

• In 2015, California and Massachusetts were the states with the highest full-attendance rates: 51.1 and 51.0 percent, respectively, of their students did not miss any school days; they are closely followed by Virginia (48.4 percent) and Illinois and Indiana (48.3 percent).
• At the other end of the spectrum, Utah and Wyoming had the largest shares of students missing more than 10 days of school in the month prior to the 2015 assessment (4.6 and 3.5 percent, respectively).
• Five states and Washington, D.C., stood out for their high shares of students missing three or more days of school in 2015: in Utah, nearly two-thirds of students (63.5 percent) missed three or more days; in Alaska, nearly half (49.6 percent) did; and in the District of Columbia, Wyoming, New Mexico, and Montana, nearly three in 10 students were in this absenteeism category.
• In most states, overall absenteeism rates changed little between 2003 and 2015.

Prior research linking chronic absenteeism with lowered academic performance is confirmed by our results. As expected, and as states have long understood, missing school is negatively associated with academic performance (after controlling for factors including race, poverty status, gender, IEP status, and ELL status). As students miss school more frequently, their performance worsens.

• Overall performance gaps. The gaps in math scores between students who did not miss any school and those who missed three or more days of school varied from 0.3 standard deviations (for students who missed 3–4 days of school the month prior to when the assessment was taken) to close to two-thirds of a standard deviation (for those who missed more than 10 days of school). The gap between students who did not miss any school and those who missed just 1–2 days of school was 0.10 standard deviations, a statistically significant but relatively small difference in practice.
• Performance gaps by groups.7 The relationship between absenteeism and performance poses problems for all students, but the degree to which absenteeism affects performance varies somewhat across student groups.
  • For Hispanic non-ELL students, missing more than 10 days of school harmed their performance on the math assessment more strongly than for the average (0.74 standard deviations vs. 0.64 on average).
  • For Asian non-ELL students, the penalty for missing school was smaller than the average (except for those missing 5–10 days).
  • Missing school hindered performance similarly across the three poverty-status groups (nonpoor, somewhat poor, and poor). However, given that there are substantial differences in the frequency with which children miss school by poverty status (that is, poor students are more likely to be chronically absent than nonpoor students), absenteeism may in fact further widen income-based achievement gaps.

What do we already know about why children miss school and which children miss school? What do we add to this evidence?

Poor health, parents’ nonstandard work schedules, low socioeconomic status (SES), changes in adult household composition (e.g., adults moving into or out of the household), residential mobility, and extensive family responsibilities (e.g., children looking after siblings)—along with inadequate supports for students within the educational system (e.g., lack of adequate transportation, unsafe conditions, lack of medical services, harsh disciplinary measures, etc.)—are all associated with a greater likelihood of being absent, and particularly with being chronically absent (Ready 2010; U.S. Department of Education 2016).8 Low-income students and families disproportionately face these challenges, and some of these challenges may be particularly acute in disadvantaged areas9; residence in a disadvantaged area may therefore amplify or reinforce the distinct negative effects of absenteeism on educational outcomes for low-income students.

A detailed 2016 report by the U.S. Department of Education showed that students with disabilities were more likely to be chronically absent than students without disabilities; Native American and Pacific Islander students were more likely to be chronically absent than students of other races and ethnicities; and non-ELL students were more likely to be chronically absent than ELL students.10 It also showed that students in high school were more likely to miss school than students in other grades, and that about 500 school districts reported that 30 percent or more of their students missed at least three weeks of school in 2013–2014 (U.S. Department of Education 2016).

Our analysis complements this evidence by adding several dimensions to the breakdown of who misses school—including absenteeism rates by poverty status and state—and by analyzing how missing school harms performance. We distinguish by the number of school days students report having missed in the month prior to the assessment (using five categories, from no days missed to more than 10 days missed over the month),11 and we compare absenteeism rates across grades and across cohorts (between 2003 and 2015), as available in the NAEP data.12

How much school are children missing? Are they missing more days than the previous generation?

In 2015, almost one in five, or 19.2 percent of, eighth-grade students missed three or more days of school in the month before they participated in NAEP testing.13 About 13 percent missed 3–4 days, roughly 5 percent missed 5–10 days, and a small share—less than 2 percent—missed more than 10 days, or half or more of the instructional days that month (Figure A, bottom panel).14

Figure A

On average, however, students in 2015 did not miss any more days than students in the earlier period; by some measures, they missed less school than children in 2003 (Figure A, top panel). While the share of students with occasional absences (1–2 days) increased moderately between 2003 and 2015, the share of students who missed more than three days of school declined by roughly 3 percentage points between 2003 and 2015. This reduction was distributed about evenly (in absolute terms) across the shares of students missing 3–4, 5–10, and more than 10 days of school. But in relative terms, the reduction was much more significant in the share of students missing more than 10 days of school (the share decreased by nearly one-third). We find no significant differences by grade (Appendix Figure A) or by subject. Thus, we have chosen to focus our analyses below on the sample of eighth-graders taking the math assessment only.

Which groups miss school most often? Which groups suffer the most from chronic absenteeism?

Absenteeism by race/ethnicity and language status

Hispanic ELLs and the group made up of Native Americans plus “all other races” (not white, black, Hispanic, or Asian) are the racial/ethnic and language status groups that missed school most frequently in 2015. Only 39.6 percent (Native American or other) and 41.2 percent (Hispanic ELL) did not miss any school in the month prior to the assessment (vs. 44.4 percent overall, 43.2 percent for white students, 43.5 percent for black students, and 44.1 percent for Hispanic non-ELL students; see Figure B1).15

Figure B1

Asian students (both non-ELL and ELL) are the least likely among all racial/ethnic student groups to be absent from school at all. Two-thirds of Asian non-ELL students and almost as many (61.6 percent of) Asian ELL students did not miss any school. Among Asian non-ELL students, only 8.8 percent missed three or more days of school: 6.1 percent missed 3–4 days (12.7 percent on average), 2.1 percent missed 5–10 days (relative to 4.8 percent for the overall average), and only 0.6 percent missed more than 10 days of school (relative to 1.7 percent for the overall average). Among Asian ELL students, the share who missed three or more days of school was 13.3 percent.

As seen in Figure B2, the differences in absenteeism rates between white students and Hispanic non-ELL students were relatively small, when looking at the shares of students missing three or more days of school (18.3 percent and 19.1 percent, respectively). The gaps are somewhat larger for black, Native American, and Hispanic ELL students relative to white students (with shares missing three or more days at 23.0, 24.0, and 24.1 percent, respectively, relative to 18.3 percent for white students).

Figure B2

Among students who missed a lot of school (more than 10 days), there were some more substantial differences by race and language status. About 3.9 percent of Hispanic ELL students and 3.2 percent of Asian ELL students missed more than 10 days of school, compared with 2.2 percent for Native American and other races, 2.0 percent for black students, 1.4 percent for white students, and only 0.6 percent for Asian non-ELL students (all relative to the overall average of 1.7 percent) (see Figure B3).

Figure B3

Absenteeism by income status

The attendance gaps are even larger by income status than they are by race/ethnicity and language status (Figures B1–B3). Poor (free-lunch-eligible) students were 5.9 percentage points more likely to miss some school than nonpoor (non-FRPL-eligible) students, and they were 7.8 percentage points more likely to miss school three or more days (23.2 vs. 15.4 percent).16 Among somewhat poor (reduced-price-lunch-eligible) students, 17.9 percent missed three or more days of school. The lowest-income (free-lunch-eligible) students were 4.1 percentage points more likely to miss school 3–4 days than non-FRPL-eligible students, and more than 2.4 percentage points more likely to miss school 5–10 days (Appendix Figure B). Finally, and most striking, free-lunch-eligible students—the most economically disadvantaged students—were more than twice as likely to be absent from school for more than 10 days as nonpoor students. In other words, they were much more likely to experience extreme chronic absenteeism. Figures B1–B3 show that the social-class gradient for the prevalence of absenteeism, proxied by eligibility for free or reduced-price lunch, is noticeable in all absenteeism categories, and especially when it comes to those students who missed the most school.

Absenteeism by disability status

Students with IEPs were by far the most likely to miss school relative to all other groups.17 The share of IEP students missing school exceeded the share of non-IEP students missing school by 7.7 percentage points (Figure B1). More than one in four IEP students had missed school three days or more in the previous month (Figure B2). About 15.5 percent of students with IEPs missed school 3–4 days (vs. 12.4 percent among non-IEP students); 7.3 percent missed 5–10 days; and 3.2 percent missed more than 10 days of school in the month before being tested (Appendix Figure B; Figure B3).

Absenteeism by gender

The differences by gender are slightly surprising (Figures B1–B3). Boys showed a higher full-attendance rate than girls (46.6 vs. 42.1 percent did not miss any school), and boys were no more likely than girls to display extreme chronic absenteeism (1.7 percent of boys and 1.6 percent of girls missed more than 10 days of school). Boys (18.2 percent) were also slightly less likely than girls (20.2 percent) to be chronically absent (to miss three or more days of school, as per our definition).

Has there been any change over time in which groups of children are most often absent from school?

For students in several groups, absenteeism fell between 2003 and 2015 (Figure C1), in keeping with the overall decline noted above. Hispanic students (both ELL and non-ELL), Asian non-ELL students, Native American and other race students, free-lunch-eligible (poor) students, reduced-priced-lunch-eligible (somewhat poor) students, non-FRPL-eligible (nonpoor) students, and IEP students were all less likely to miss school in 2015 than they were over a decade earlier. For non-IEP and white students, however, the share of students who did not miss any school days in the month prior to NAEP testing remained essentially unchanged, while it increased slightly for black students and Asian ELL students (by about 2 percentage points each).

Figure C1

As seen in Figure C2, we also note across-the-board reductions in the shares of students who missed three or more days of school (with the exception of the share of Asian ELL students, which increased by 1.7 percentage points over the time studied). The largest reductions occurred for students with disabilities (IEP students), Hispanic non-ELL students, Native American students or students of other races, free-lunch-eligible students, and non-FRPL-eligible students (each of these groups experienced a reduction of at least 4.4 percentage points).18 For all groups except Asian ELL students, the share of students missing more than 10 days of school (Figure C3) also decreased (for Asian ELL students, it increased by 1.3 percentage points).

Figure C2

Figure C3

In order to get a full understanding of these comparisons, we need to look at both the absolute and relative differences. Overall, the data presented show modest absolute differences in the shares of students who are absent (at any level) in various groups when compared with the averages for all students (Figures B1–B3 and Appendix Figure B). The differences (both absolute and relative) among student groups missing a small amount of school (1–2 days) are minimal for most groups. However, while the differences among groups are very small in absolute terms for students missing a lot of school (more than 10 days), some of the differences are very large in relative terms. (And, taking into account the censoring problem mentioned earlier, they could potentially be even larger.)

The fact that the absolute differences are small is in marked contrast to differences seen in many other education indicators of outcomes and inputs, which tend to be much larger by race and income divisions (Carnoy and García 2017; García and Weiss 2017). Nevertheless, both the absolute and relative differences we find are revealing and important, and they add to the set of opportunity gaps that harm students’ performance.

Is absenteeism particularly high in certain states?

In 2015, California and Massachusetts had the highest full-attendance rates among the states (51.1 and 51.0 percent of students, respectively, did not miss any school days), closely followed by Virginia (48.4 percent) and Illinois and Indiana (48.3 percent), while Alaska, the District of Columbia, Montana, New Mexico, and Utah had the lowest rates of perfect attendance (fewer than one in three students had perfect attendance). The latter group of states also had the highest rates of chronic or extreme chronic absenteeism: Utah had by far the largest share of students missing school three or more days (63.5 percent), followed by Alaska (49.6 percent), and Wyoming, New Mexico, Montana, and the District of Columbia (the latter four ranging from 27.5 to 29.8 percent). Utah also had the largest share of students missing school more than 10 days (4.6 percent), followed by Wyoming, Montana, and the District of Columbia (3.5, 3.3, and 3.2 percent, respectively). (See the Interactive Map for data for all states.)

Interactive Map

Over the 2003–2015 period, 22 states saw their share of students with perfect attendance grow. The number drops to 15 if we count only states in which the share of students not missing any school increased by more than 1 percentage point. In almost every state (44 states), the share of students who missed more than 10 school days decreased, and in 41 states, the share of students who missed three or more days of school also dropped, though it increased in the other 10.19 Louisiana, Massachusetts, Nevada, Indiana, New Hampshire, and California were the states in which these shares decreased the most, by more than 6 percentage points, while Utah, Alaska, and North Dakota were the states where this indicator (three or more days missed) showed the worst trajectory over time (that is, the largest increases in chronic absenteeism).

Is absenteeism a problem for student performance?

Previous research has focused mainly on two groups of students when estimating how much absenteeism influences performance: students who are chronically absent and all other students. This prior research has concluded that students who are chronically absent are at serious risk of falling behind in school, having lower grades and test scores, having behavioral issues, and, ultimately, dropping out (U.S. Department of Education 2016; see summary in Gottfried and Ehrlich 2018). Our analysis allows for a closer examination of the relationship between absenteeism and performance, as we look at the impact of absenteeism on student performance at five levels of absenteeism. This design allows us to test not only whether different levels of absenteeism have different impacts on performance (as measured by NAEP test scores), but also to identify the point at which the impact of absenteeism on performance becomes a concern. Specifically, we look at the relationship between student absenteeism and mathematics performance among eighth-graders at various numbers of school days missed.20

The results shown in Figure D and Appendix Table 1 are obtained from regressions that assess the influence of absenteeism and other individual- and school-level determinants of performance. The latter include students’ race/ethnicity, gender, poverty status, ELL status, and IEP status, as well as the racial/ethnic composition of the school they attend and the share of students in their school who are eligible for FRPL (a proxy for the SES composition of the school). Our results thus identify the distinct association between absenteeism and performance, net of other factors that are known to influence performance.21

In general, the more frequently children missed school, the worse their performance. Relative to students who didn’t miss any school, those who missed some school (1–2 school days) accrued, on average, an educationally small, though statistically significant, disadvantage of about 0.10 standard deviations (SD) in math scores (Figure D and Appendix Table 1, first row). Students who missed more school experienced much larger declines in performance. Those who missed 3–4 days or 5–10 days scored, respectively, 0.29 and 0.39 standard deviations below students who missed no school. As expected, the harm to performance was much greater for students who were absent half or more of the month. Students who missed more than 10 days of school scored nearly two-thirds (0.64) of a standard deviation below students who did not miss any school. All of the gaps are statistically significant, and together they identify a structural source of academic disadvantage.

Figure D

The results show that missing school has a negative effect on performance regardless of how many days are missed, with a moderate dent in performance for those missing 1–2 days and a troubling decline in performance for students who missed three or more days that becomes steeper as the number of missed days rises to 10 and beyond. The point at which the impact of absenteeism on performance becomes a concern, therefore, is when students miss any amount of school (vs. having perfect attendance); the level of concern grows as the number of missed days increases.

Gaps in performance associated with absenteeism are similar across all races/ethnicities, between boys and girls, between FRPL-eligible and noneligible students, and between students with and without IEPs. For example, relative to nonpoor (non-FRPL-eligible) students who did not miss any school, nonpoor children who missed school accrued a disadvantage of -0.09 SD (1–2 school days missed), -0.27 SD (3–4 school days missed), -0.36 SD (5–10 school days missed), and -0.63 SD (more than 10 days missed). For students eligible for reduced-price lunch (somewhat poor students) who missed school, compared with students eligible for reduced-price lunch who did not miss any school, the gaps are -0.16 SD (1–2 school days missed), -0.33 SD (3–4 school days missed), -0.45 SD (5–10 school days missed), and -0.76 SD (more than 10 days missed). For free-lunch-eligible (poor) students who missed school, relative to poor students who do not miss any school, the gaps are -0.11 SD (1–2 school days missed), -0.29 SD (3–4 school days missed), -0.39 SD (5–10 school days missed), and -0.63 SD (more than 10 days missed). By IEP status, relative to non-IEP students who did not miss any school, non-IEP students who missed school accrued a disadvantage of -0.11 SD (1–2 school days missed), -0.30 SD (3–4 school days missed), -0.40 SD (5–10 school days missed), and -0.66 SD (more than 10 days missed). And relative to IEP students who did not miss any school, IEP students who missed school accrued a disadvantage of -0.05 SD (1–2 school days missed), -0.21 SD (3–4 school days missed), -0.31 SD (5–10 school days missed), and -0.52 SD (more than 10 days missed). (For gaps by gender and by race/ethnicity, see Appendix Table 1).

Importantly, though the gradients of the influence of absenteeism on performance by race, poverty status, gender, and IEP status (Appendix Table 1) are generally similar to the gradients in the overall relationship between absenteeism and performance for all students, this does not mean that all groups of students are similarly disadvantaged when it comes to the full influence of absenteeism on performance. The overall performance disadvantage faced by any given group is influenced by multiple factors, including the size of the group’s gaps at each level of absenteeism (Appendix Table 1), the group’s rates of absenteeism (Figure B), and the relative performance of the group with respect to the other groups (Carnoy and García 2017). The total gap that results from adding these factors can thus become substantial.

To illustrate this, we look at Hispanic ELL, Asian non-ELL, Asian ELL, and FRPL-eligible students. The additional penalty associated with higher levels of absenteeism is smaller than average for Hispanic ELL students experiencing extreme chronic absenteeism; however, their performance is the lowest among all groups (Carnoy and García 2017) and they have among the highest absenteeism rates.

The absenteeism penalty is also smaller than average for Asian non-ELL students (except at 5-10 days); however, in contrast with the previous example, their performance is the highest among all groups (Carnoy and García 2017) and their absenteeism rate is the lowest.

The absenteeism penalty for Asian ELL students is larger than average, and the gradient is steeper.22 Asian ELL students also have lower performance than most other groups (Carnoy and García 2017).

Finally, although there is essentially no difference in the absenteeism–performance relationship by FRPL eligibility, the higher rates of absenteeism (at every level) for students eligible for free or reduced-price lunch, relative to nonpoor (FRPL-ineligible) students, put low-income students at a greater risk of diminished performance due to absenteeism than their higher-income peers, widening the performance gap between these two groups.

Conclusions

Student absenteeism is a puzzle composed of multiple pieces that has a significant influence on education outcomes, including graduation and the probability of dropping out. The factors that contribute to it are complex and multifaceted, and likely vary from one school setting, district, and state to another. This analysis aims to shed additional light on some key features of absenteeism, including which students tend to miss school, how those profiles have changed over time, and how much missing school matters for performance.

Our results indicate that absenteeism rates were high and persistent over the period examined (2003–2015), although they did decrease modestly for most groups and in most states. Unlike findings for other factors that drive achievement gaps—from preschool attendance to economic and racial school segregation to unequal funding (Carnoy and García 2017; García 2015; García and Weiss 2017)—our findings here seem to show some positive news for black and Hispanic students: these students had slightly higher perfect attendance rates than their white peers; in addition, their perfect attendance rates have increased over time at least as much as rates for white students. But with respect to the absenteeism rates that matter the most (three or more days of school missed, and more than 10 days of school missed), black and Hispanic students still did worse (just as is the case with other opportunity gaps faced by these students). Particularly worrisome is the high share of Hispanic ELL students who missed more than 10 school days—nearly 4 percent. Combined with the share of Hispanic ELL students who missed 5–10 school days (nearly 6 percent), this suggests that one in 10 children in this group would miss school for at least a quarter of the instructional time.

The advantages that Asian students enjoy relative to white students and other racial/ethnic groups in academic settings is also confirmed here (especially among Asian non-ELL students): the Asian students in the sample missed the least school. And there is a substantial difference in rates of absenteeism by poverty (FRPL) and disability (IEP) status, with the difference growing as the number of school days missed increases. Students who were eligible for free lunch were twice as likely as nonpoor (FRPL-ineligible) students to be absent more than 10 days, and students with IEPs were more likely than any other group to be absent (one or more days, that is, to not have perfect attendance).

Missing school has a distinct negative influence on performance, even after the potential mediating influence of other factors is taken into account, and this is true at all rates of absenteeism. The bottom line is that the more days of school a student misses, the poorer his or her performance will be, irrespective of gender, race, ethnicity, disability, or poverty status.

These findings help establish the basis for an expanded analysis of absenteeism along two main, and related, lines of inquiry. One, given the marked and persistent patterns of school absenteeism, it is important to continue to explore and document why children miss school—to identify the full set of factors inside and outside of schools that influence absenteeism. Knowing whether (or to what degree) those absences are attributable to family circumstances, health, school-related factors, weather, or other factors, is critical to effectively designing and implementing policies and practices to reduce absenteeism, especially among students who chronically miss school. The second line of research could look at variations in the prevalence and influence of absenteeism among the states, and any changes over time in absenteeism rates within each state, to assess whether state differences in policy are reducing absenteeism and mitigating its negative impacts. For example, in recent years, Connecticut has made reducing absenteeism, especially chronic absenteeism, a top education policy priority, and has developed a set of strategies and resources that could be relevant to other states as well, especially as they begin to assess and respond to absenteeism as part of their ESSA plans.23

The analyses in this report confirm the importance of looking closely into “other” education data, above and beyond performance (test scores) and individual and school demographic characteristics. The move in education policy toward widening accountability indicators to indicators of school quality, such as absenteeism, is important and useful, and could be expanded to include other similar data. Indicators of bullying, school safety, student tardiness, truancy, level of parental involvement, and other factors that are relevant to school climate, well-being, and student performance would also merit attention.

Acknowledgements

The authors gratefully acknowledge John Schmitt and Richard Rothstein for their insightful comments and advice on earlier drafts of the paper. We are also grateful to Krista Faries for editing this report, to Lora Engdahl for her help structuring it, and to Julia Wolfe for her work preparing the tables and figures included in the appendix. Finally, we appreciate the assistance of communications staff at the Economic Policy Institute who helped to disseminate the study, especially Dan Crawford and Kayla Blado.

About the authors

Emma García is an education economist at the Economic Policy Institute, where she specializes in the economics of education and education policy. Her areas of research include analysis of the production of education, returns to education, program evaluation, international comparative education, human development, and cost-effectiveness and cost-benefit analysis in education. Prior to joining EPI, García was a researcher at the Center for Benefit-Cost Studies of Education, the National Center for the Study of Privatization in Education, and the Community College Research Center at Teachers College, Columbia University, and did consulting work for the National Institute for Early Education Research, MDRC, and the Inter-American Development Bank. García has a Ph.D. in economics and education from Teachers College, Columbia University.

Elaine Weiss served as the national coordinator for the Broader, Bolder Approach to Education (BBA) from 2011 to 2017, in which capacity she worked with four co-chairs, a high-level task force, and multiple coalition partners to promote a comprehensive, evidence-based set of policies to allow all children to thrive. She is currently working on a book drawing on her BBA case studies, co-authored with Paul Reville, to be published by the Harvard Education Press. Weiss came to BBA from the Pew Charitable Trusts, where she served as project manager for Pew’s Partnership for America’s Economic Success campaign. Weiss was previously a member of the Centers for Disease Control and Prevention’s task force on child abuse and served as volunteer counsel for clients at the Washington Legal Clinic for the Homeless. She holds a Ph.D. in public policy from the George Washington University and a J.D. from Harvard Law School.

Appendix figures and tables

Appendix Figure A

Appendix Figure B

Appendix Table 1

Endnotes

1. See García 2014 and García and Weiss 2016.

2. See ESSA 2015. According to ESSA, this nontraditional indicator should measure “school quality or student success.” (The other indicators at elementary/middle school include measures of academic achievement, e.g., performance or proficiency in reading/language arts and math; academic progress, or student growth; and progress in achieving English language proficiency.)

3. Thirty-six states and the District of Columbia have included student absenteeism as an accountability metric in their states’ ESSA plans. This metric meets all the requirements (as outlined in ESSA) to be considered a measure of school quality or student success (valid, reliable, calculated the same for all schools and school districts across the state, can be disaggregated by student subpopulation, is a proven indicator of school quality, and is a proven indicator of student success; see Education Week 2017). See FutureEd 2017 for differences among the states’ ESSA plans. See the web page “ESSA Consolidated State Plans” (on the Department of Education website) for the most up-to-date information on the status and content of the state plans.

4. There is no precise official definition that identifies how many missed days constitutes chronic absenteeism on a monthly basis. Definitions of chronic absenteeism are typically based on the number of days missed over an entire school year, and even these definitions vary. For the Department of Education, chronically absent students are those who “miss at least 15 days of school in a year” (U.S. Department of Education 2016). Elsewhere, chronic absenteeism is frequently defined as missing 10 percent or more of the total number of days the student is enrolled in school, or a month or more of school, in the previous year (Ehrlich et al. 2013; Balfanz and Byrnes 2012). Given that the school year can range in length from 180 to 220 days, and given that there are about 20–22 instructional days in a month of school, these latter two definitions imply that a student is chronically absent if he or she misses between 18 and 22 days per year (depending on the length of the school year) or more, or between 2.0 and about 2.5 days (or more) per month on average (assuming a nine-month school year). In our analysis, we define students as being chronically absent if they have missed three or more days of school in the last month (the aggregate of students missing “3–4,” “5–10,” or “more than 10 days”), and as experiencing extreme chronic absenteeism if they have missed “more than 10 days” of school in the last month. These categories are not directly comparable to categories used in studies of absenteeism on a per-year basis or that use alternative definitions or thresholds. We purposely analyze data for each of these “days absent” groups separately to identify their distinct characteristics and the influence of those differences on performance. (Appendix Figure B and Appendix Table 1 provide separate results for each of the absenteeism categories.)

5.  In our analysis, we define “poor” students as those who are eligible for free lunch; we define “somewhat poor” students as those who are eligible for reduced-price lunch; and we define “nonpoor” students as those who are not eligible for free or reduced-price lunch. We use “poverty status,” “income status,” “socioeconomic status” (“SES”), and “social class” interchangeably throughout our analysis. We use the free or reduced-price lunch status classification as a metric for individual poverty, and we use the proportion of students who are eligible for FRPL as a metric for school poverty (in our regression controls; see Figure D). The limitations of these variables to measure economic status are discussed in depth in Michelmore and Dynarski’s (2016) study. FRPL statuses are nevertheless valid and widely used proxies of low(er) SES, and students’ test scores are likely to reflect such disadvantage (Carnoy and García 2017).

6. Under the Individuals with Disabilities Education Act (IDEA), an IEP must be designed for each student with a disability. The IEP “guides the delivery of special education supports and services for the student” (U.S. Department of Education 2000). For more information about IDEA, see U.S. Department of Education n.d.

7. Students are grouped by gender, race/ethnicity and ELL status, FRPL eligibility, and IEP status.

8. The U.S. Department of Education (2016) defines “chronically absent” as “missing at least 15 days of school in a year.” Ready (2010) explains the difference between legitimate or illegitimate absences, which may respond to different circumstances and behaviors. Ready’s findings, pertaining to children at the beginning of school, indicate that, relative to high-SES students, low-SES children with good attendance rates experienced greater gains in literacy skills during kindergarten and first grade, narrowing the starting gaps with their high-SES peers. No differences in math skills gains were detected in kindergarten.

9. U.S. Department of Education 2016. This report uses data from the Department of Education’s Civil Rights Data Collection 2013–2014.

10. The analysis finds no differences in absenteeism by gender. It is notable that the Department of Education report finds that ELL students have lower absenteeism rates than their non-ELL peers, given that we find (as described later in the report) that Asian ELL students have higher absenteeism rates than Asian non-ELL students and that Hispanic ELL students have higher absenteeism rates than Hispanic non-ELL students. It is important to note, however, that the data the Department of Education analyze compared all ELL students to all non-ELL students (not only Asian and Hispanic students separated out by ELL status), and thus our estimates are not directly comparable.

11. Children in the fourth and eighth grades were asked, “How many days were you absent from school in the last month?” The possible answers are: none, 1–2 days, 3–4 days, 5–10 days, and more than 10 days. An important caveat concerning this indicator and results based on its utilization is that there is a potential inherent censoring problem: Children who are more likely to miss school are also likely to miss the assessment. In addition, some students may be inclined to underreport the number of days that they missed school, in an effort to be viewed more favorably (in social science research, this may introduce a source of response-bias referred to as “social desirability bias”). Although we do not have any way to ascertain the extent to which these might be problems in the NAEP data and for this question in particular, it is important to read our results and findings as a potential underestimate of what the rates of missingness are, as well as what their influence on performance is.

12. One reason to look at different grades is to explore the potential connection between early absenteeism and later absenteeism. Ideally, we would be able to include data on absenteeism from earlier grades in students’ academic careers since, as Nai-Lin Chang, Sundius, and Wiener (2017) explain, attendance habits are developed early and often set the stage for attendance patterns later on. These authors argue that detecting absenteeism early on can improve pre-K to K transitions, especially for low-income children, children with special needs, or children who experience other challenges at home; these are the students who most need the social, emotional, and academic supports that schools provide and whose skills are most likely to be negatively influenced by missing school. Gottfried (2014) finds reduced reading and math achievement outcomes, and lower educational and social engagement, among kindergartners who are chronically absent. Even though we do not have information on students’ attendance patterns at the earliest grades, looking at patterns in the fourth and eighth grades can be illuminating.

13. Students are excluded from our analyses if their absenteeism information and/or basic descriptive information (gender, race/ethnicity, poverty status, and IEP) are missing.

14. All categories combined, we note that in 2015, 49.5 percent of fourth-graders and 55.6 percent of eighth-graders missed at least one day of school in the month prior. Just over 30 percent of fourth-graders and 36.4 percent of eighth-graders missed 1–2 days of school during the month.

15. In the sample, 52.1 percent of students are white, 14.9 percent black, 4.5 percent Hispanic ELL, 19.4 percent Hispanic non-ELL, less than 1 percent Asian ELL, 4.7 percent Asian non-ELL, and 3.8 percent Native American or other.

16. Of the students in the sample, 47.8 percent are not eligible for FRPL, 5.2 percent are eligible for reduced-price lunch, and 47.0 percent are eligible for free lunch.

17. In the 2015 eighth-grade mathematics sample, 10.8 percent of students had an IEP.

18. For students who were eligible for reduced-price lunch (somewhat poor students), shares of students absent three or more days also decreased, but more modestly, by 3.3 percentage points.

19. Number of states is out of 51; the District of Columbia is included in the state data.

20. The results discussed below cannot be interpreted as causal, strictly speaking. They are obtained using regression models with controls for the relationship between performance and absenteeism (estimates are net of individual, home, and school factors known to influence performance and are potential sources of selection). However, the literature acknowledges a causal relationship between (high-quality) instructional time and performance, in discussions about the length of the school day (Kidronl and Lindsay 2014; Jin Jez and Wassmer 2013; among others) and the dip in performance children experience after being out of school for the summer (Peterson 2013, among others). These findings could be extrapolable to our absenteeism framework and support a more causal interpretation of the findings of this paper.

21. Observations with full information are used in the regressions. The absenteeism–performance relationship is only somewhat sensitive to including traditional covariates in the regression (not shown in the tables; results available upon request). The influence of absenteeism on performance is distinct and is not due to any mediating effect of the covariates that determine education performance.

22. Asian ELL students who miss more than 10 days of school are very far behind Asian ELL students with perfect attendance, with a gap of more than a standard deviation. This result needs to be interpreted with caution, however, as it is based on a very small fraction of students for whom selection may be a concern, too.

23. The data used in our analysis are for years prior to the implementation of measures intended to tackle absenteeism. See Education Week 2017. Data for future (or more recent) years will be required to analyze whether Connecticut’s policies have had an effect on absenteeism rates in the state.

References

Balfanz, Robert, and Vaughan Byrnes. 2012. The Importance of Being in School: A Report on Absenteeism in the Nation’s Public Schools. Johns Hopkins University Center for Social Organization of Schools, May 2012.

Carnoy, Martin, and Emma García. 2017. Five Key Trends in U.S. Student Performance: Progress by Blacks and Hispanics, the Takeoff of Asians, the Stall of Non-English Speakers, the Persistence of Socioeconomic Gaps, and the Damaging Effect of Highly Segregated Schools. Economic Policy Institute, January 2017.

Education Week. 2017. School Accountability, School Quality and Absenteeism under ESSA (Expert Presenters: Hedy Chang and Charlene Russell-Tucker) (webinar).

Ehrlich, Stacy B., Julia A. Gwynne, Amber Stitziel Pareja, and Elaine M. Allensworth with Paul Moore, Sanja Jagesic, and Elizabeth Sorice. 2013. Preschool Attendance in Chicago Public Schools: Relationships with Learning Outcomes and Reasons for Absences. The University of Chicago Consortium on Chicago School Research, September 2013.

ESSA. 2015. Every Student Succeeds Act of 2015, Pub. L. No. 114-95 § 114 Stat. 1177 (2015–2016).

FutureEd. 2017. Chronic Absenteeism and the Fifth Indicator in State ESSA Plans. Georgetown University.

García, Emma. 2014. The Need to Address Noncognitive Skills in the Education Policy Agenda. Economic Policy Institute, December 2014.

García, Emma. 2015. Inequalities at the Starting Gate: Cognitive and Noncognitive Skills Gaps between 2010–2011 Kindergarten Classmates. Economic Policy Institute, June 2015.

García, Emma, and Elaine Weiss. 2016. Making Whole-Child Education the Norm. How Research and Policy Initiatives Can Make Social and Emotional Skills a Focal Point of Children’s Education. Economic Policy Institute, August 2016.

García, Emma, and Elaine Weiss. 2017. Education Inequalities at the School Starting Gate: Gaps, Trends, and Strategies to Address Them. Economic Policy Institute, September 2017.

Gottfried, Michael A. 2014. “Chronic Absenteeism and Its Effects on Students’ Academic and Socioemotional Outcomes.” Journal of Education for Students Placed at Risk 19, no. 2: 53–75. https://doi.org/10.1080/10824669.2014.962696.

Gottfried, Michael A., and Stacy B. Ehrlich. 2018. “Introduction to the Special Issue: Combating Chronic Absence.” Journal of Education for Students Placed at Risk 23, no. 1–2: 1–4. https://doi.org/10.1080/10824669.2018.1439753.

Jin Jez, Su, and Robert W. Wassmer. 2013. “The Impact of Learning Time on Academic Achievement.” Education and Urban Society 47, no. 3: 284–306. https://doi.org/10.1177/0013124513495275.

Kidronl, Yael, and Jim Lindsay. 2014. The Effects of Increased Learning Time on Student Academic and Nonacademic Outcomes: Findings from a Meta-Analytic Review. REL 2014-015. Regional Educational Laboratory Appalachia.

Michelmore, K., and S. Dynarski. 2016. The Gap within the Gap: Using Longitudinal Data to Understand Income Differences in Student Achievement. National Bureau of Economic Research Working Paper no. 22474.

Nai-Lin Chang, Hedy, Jane Sundius, and Louise Wiener. 2017. “Using ESSA to Tackle Chronic Absence from Pre-K to K–12” (blog post). National Institute for Early Education Research website, May 23, 2017.

National Center for Education Statistics (NCES), National Assessment of Educational Progress (NAEP). Various years. NAEP microdata (unpublished data).

Peterson, T.K., ed. 2013. Expanding Minds and Opportunities: Leveraging the Power of Afterschool and Summer Learning for Student Success. Washington, D.C.: Collaborative Communications Group.

Ready, Douglas D. 2010. “Socioeconomic Disadvantage, School Attendance, and Early Cognitive Development: The Differential Effects of School Exposure.” Sociology of Education 83, no. 4: 271–286. https://doi.org/10.1177/0038040710383520.

U.S. Department of Education. 2000. A Guide to the Individualized Education Program. Office of Special Education and Rehabilitative Services, July 2000.

U.S. Department of Education. 2016. Chronic Absenteeism in the Nation’s Schools: An Unprecedented Look at a Hidden Educational Crisis (online fact sheet).

U.S. Department of Education. n.d. “About IDEA” (webpage). IDEA (Individuals with Disabilities Education Act) website. Accessed September 19, 2018.

The American Institutes of Research (AIR) has a new report by Michelle Yin and Dahlia Shaewitz showing that the labor force participation of Americans with disabilities fell from 25 percent in 2001 to 16 percent in 2014, based on data from the Current Population Survey Annual Social and Economic Supplement conducted by the U.S. Census Bureau (CPS-ASEC—henceforth CPS). Tying this to a broader decline in the labor force participation of working-age adults, the authors warn that “this situation leaves the United States with an even smaller pool of workers to support the recovering economy. “

In the same vein, a recent op-ed in the Wall Street Journal by Andrew Biggs of the American Enterprise Institute cites a “nearly 50 percent decline in the employment rate of Americans with disabilities since 1981.” Echoing critiques of the Social Security Disability Insurance (SSDI) program I’ve discussed in earlier blog posts, Biggs attributes the problem  to “looser eligibility standards and stagnating wages that made disability benefits, averaging $1,222 a month for new beneficiaries last year, more attractive relative to work for the less-skilled.” Though Yin and Shaewitz appear more concerned with the plight of people with disabilities than with criticizing SSDI, they also suggest that the “growing number of discouraged workers with disabilities may be a result of policies that unintentionally make it easier to leave the workforce or stay out altogether.”

Are employment rates declining because people with disabilities are leaving the workforce? More likely, it’s the other way around—labor market weakness has affected all workers, regardless of disability status. Yin and Shaewitz discount the effects of a weak labor market, noting that “U.S. unemployment rates are nearly back to normal after the Great Recession.” But although unemployment may appear to be back to normal, employment is not, due to a large number of discouraged workers who aren’t actively looking and therefore aren’t officially counted as unemployed. In fact, employment rates never fully recovered from the effects of the 2001 recession, let alone the much deeper 2007-2009 recession. (Unless otherwise noted, all statistics are based on an EPI analysis of CPS data.)

If workers with disabilities were leaving the workforce in larger numbers than healthy workers for reasons such as easier access to benefits or greater financial incentives for applying, we would expect them to be a growing share of the non-working population. This is generally not what we find. Men with disabilities are a stable or declining share of non-working men, as shown in the chart below. The exception is men in their 60s—but this is because fewer men in their 60s are retiring, not because more are leaving the workforce because of a disability.

Disability and Labor Force Participation

A similar pattern is observed for women. For the most part, women with disabilities are a stable share of non-working women, though there has been an increase in the disabled share of non-working women in their 50s and 60s. This increase may be because of the rise in women’s labor force participation, as well as the trend toward later retirement; since more women in their 50s and 60s are in the workforce, there are more women to report a work-limiting disability.

In 2000, 42 percent of women in their 50s who weren’t in the paid workforce cited caring for home or family as the reason. This share declined to 33 percent in 2014. As the pool of non-workers shrinks and the share of women who cite caregiving for the reason they aren’t in the workforce declines, the share of women who cite disability as the reason for not working will go up. Indeed, the decline in full-time caregiving explains more than a third of the increase in the disability share for that age group.

Disability and Labor Force Participation

Admittedly, even if the employment decline is limited to older women with disabilities, older adults are the bulk of the disabled population. So it could still be problem if more older women are exiting the workforce to claim disability benefits. However, as noted in an earlier blog post, Social Security administrative data show no upward trend in age- and unemployment-adjusted disability take-up over the past 20 years, as a modest decline for men has offset a modest increase for women. Moreover, the increase for older women is due in large part to trends in labor force participation, as, over time, more women in this age group had worked long enough to qualify for benefits. CPS data generally corroborate administrative data on disability receipt.

What explains these seemingly contradictory trends? Has a weak job market had a stronger-than-usual effect on workers with disabilities? Are adults with disabilities leaving the workforce for other reasons? If so, why isn’t that reflected in SSDI take-up and other statistics? There are a number of possible explanations, none of which suggests that more workers are leaving the labor force to claim disability benefits—at least not in the past 20 years. One possibility is simply that a growing number of workers with disabilities have found themselves with neither jobs nor disability benefits. As noted in an earlier blog post, research focusing on marginal disability applicants who might be accepted or rejected by different examiners shows that a large majority—whether accepted or rejected—are never able to earn a living again.

Though this is a real concern, it’s not clear that it’s a growing problem, properly measured. The statistics cited by Yin, Shaewitz, Biggs, and others don’t take age distribution into account—a significant omission in a period when the large baby boomer cohort entered their peak disability years. The aging of the baby boomers caused an increase in the share of adults with disabilities as well as a decline in the employment rate of workers with disabilities. Younger people with disabilities are more likely to be employed than their older counterparts, many of whom exit the labor force at the onset of disability. Between 2000 and 2014, 21% of Americans with disabilities age 25-49 were employed, compared to 14% of those aged 50-64.

Another explanation is the subjective nature of self-reported disabilities, as opposed to disabilities that meet SSDI’s eligibility standards. The CPS simply asks whether respondents have “a health problem or a disability which prevents him/her from working or which limits the kind or amount of work.” Because the definition is broad, roughly twice as many people say they have a work limiting disability in the CPS as receive disability benefits (SSDI or SSI). These issues led the National Council on Disability to discourage the use of CPS data for assessing employment rates among people with disabilities, expressing concern that it could lead to ‘‘ineffective or even dangerous public policy decisions” (for a discussion, see Silverstein, Julnes and Nolan (2005)).

Along with Massachusetts Institute of Technology economist David Autor and Stanford economist Mark Duggan, Cornell economist Richard Burkhauser is a prominent expounder of the idea that employment rates for Americans with disabilities are declining. Burkhauser and various coauthors, who have considered the issue in more detail than others who hold similar views, acknowledge the limitations of CPS and other survey data—though, like them, Burkhauser fails to account for changes in the age distribution of the population.

One of the problems with self-reported disabilities acknowledged by Burkhauser is that “those reporting a work-activity limitation are less likely to be working, because those who are not working will be more likely to report that their impairment also affects their ability to work.” Thus, for example, an unemployed waitress suffering from arthritis might describe herself as having a disability after an unsuccessful job hunt, but not before.

As a result, the employment rate of people with self-reported disabilities is highly sensitive to labor market conditions, since the denominator (the number of adults with disabilities) tends to move in the opposite direction as the numerator (the number of adults with disabilities who are employed). As noted earlier, we have experienced a prolonged period of weak labor demand. In such circumstances, if even if a small number of people change their response to a survey question about work-limiting disabilities, it can have a large impact on the employment rate of adults with disabilities, who are a small share of the adult population.

Assume, for example, that eight out of 100 adults describe themselves as having a work-limiting disability, and two of them are employed (this is roughly in line with CPS data for 2000). If a single “healthy” person loses his or her job and now reports a work-limiting disability, this reduces the employment rate of people with disabilities by 11 percent (from 2/8 to 2/9). Because this is more likely to happen when employment is declining, it magnifies the apparent impact of weak labor demand on the small disabled population.

Similarly, a small reduction in the number of employed workers who report having a disability can have a large negative effect on the employment rate of people with disabilities. Using the same example as earlier, if an employed person changes his or her status from disabled to healthy, it reduces the employment rate of people with disabilities by 43 percent (from 2/8 to 1/7). It is likely that baby boomers who shun labels like “senior” are less likely to identify themselves as “disabled” than previous generations.

A tendency for workers sidelined by a weak economy to downgrade their health status, combined with a growing unwillingness by employed workers to identify as “disabled,” could be pulling down the estimated employment rate of people with disabilities. This could theoretically happen even in the absence of any decline in the employment of people previously identified as disabled, though in practice it will tend to happen when this is occurring as well, magnifying the apparent impact of a weak labor market on workers with disabilities. It could also occur without any change in measures of disability prevalence, since the two trends have offsetting effects on prevalence.

Many articles and even books have been written about problems associated with disability statistics as well as the effects of aging, women’s labor force participation, and similar factors on these measures. Many of these issues were raised, for example, in a debate over the employment effects of the Americans with Disabilities Act. While it can be difficult to disentangle these effects, the fact that neither the AIR report nor Biggs’ op-ed even allude to these challenges suggests that neither should be taken seriously.

Earlier blog posts in this series questioned Stanford economist Mark Duggan’s Senate testimony that disability incidence is rising because low-wage workers have an increasing incentive to apply for benefits instead of working or looking for work. This post considers a related claim made by Duggan: that an increase in the share of beneficiaries with mental health disorders and musculoskeletal conditions reflects “the liberalization of the program’s medical eligibility criteria that occurred in the mid-1980s” and is problematic because “the employment potential of SSDI applicants with these more subjective conditions is substantial and it is often difficult to verify the severity of these conditions.”

The first part of Duggan’s claim is true, but somewhat misleading with respect to mental health. After Congress expanded and clarified eligibility criteria in 1984 in response to an ill-conceived effort by the Reagan Administration to reduce disability rolls, there was a rebound in disabled worker awards for mental health and other conditions. But the share of awards for mental health conditions declined as the baby boomers approached retirement because mental health problems don’t increase with age as much as other conditions.

On the other hand, the share of awards due to musculoskeletal conditions has grown steadily, more than offsetting declines in the shares due to mental health, cancer, and cardiovascular disease (see Figure 1). As discussed in earlier blog posts, there has been no upward trend in overall disability incidence over the past 20 years. Moreover, the employment potential of disability applicants is very low, even among those denied benefits. Nevertheless, it is reasonable to ask why there has been an increase in the share of awards for musculoskeletal conditions, and why increases in life expectancy due to such factors as a steep decline in smoking and advances in the treatment of cardiovascular disease haven’t led to a decline in overall disability, especially since fewer jobs now require hard physical labor.

Figure 1

As will be discussed in this blog post, there are a number of explanations that have nothing to do with relaxed medical standards. A major factor is the aging of the large baby boomer cohort, because musculoskeletal ailments increase with age. An overview by Social Security Administration (SSA) actuary Tim Zayatz (2011) noted that demographic factors explained much of the increase in musculoskeletal awards, but did not provide detailed breakdowns.

Using publicly available data that doesn’t allow for precise age adjustment, the number of musculoskeletal awards by broad age group has trended upward (Figure 2), with cyclical fluctuations and a jump in 1995 that Zayatz attributes to more accurate reporting. Before 1995, SSA reported the distribution of awards based on applicants who were admitted in the initial stage. In 1995, SSA began reporting the actual diagnosis for awards allowed on appeal, which included a larger share of awards for musculoskeletal conditions.

Figure 2

Another reason for the increase in musculoskeletal conditions is that many people in poor health, including poor mental health, suffer from multiple impairments, so advances in the treatment of cardiovascular and other diseases can change the primary diagnosis without reducing overall incidence. Such medical advances also increase the share of awards for musculoskeletal diseases even if they reduce overall disability incidence. Similarly, stricter eligibility standards—disallowing awards based solely on drug and alcohol abuse or obesity (in 1997 and 1999 respectively)—likely increased the share of awards for musculoskeletal conditions.

Declining health for some groups of workers also appears to be a factor—though the reasons are not fully understood. While average life expectancy is slowly increasing, health disparities by socioeconomic status have widened considerably in recent decades, with lower-income and less-educated Americans seeing little or no increase in life expectancy. In general, it is not a good idea to make assumptions about the health of disabled beneficiaries from statistics about the general population, since disabled beneficiaries have much higher mortality rates.

A likely factor in widening health disparities and the growth in musculoskeletal disorders is obesity, which increased across demographic and socioeconomic groups, but most among less-educated women. While awards based solely on obesity are no longer allowed, obesity contributes to musculoskeletal disorders and other impairments. It is worth noting that while both obesity and smoking are linked to socioeconomic status, there is no simple behavioral explanation for growing health disparities. Obesity is as common among men with some college education as among women lacking a high-school degree. And while more working-age adults without a college degree smoke, the least and most educated groups have seen the largest declines (with high-school graduates and adults with some college education but no bachelor’s degree seeing smaller declines).

It is as problematic to generalize about employment as it is about health. Jobs in the expanding service sector often require standing, lifting, and other sustained or repetitive activity even if they don’t require as much hard physical labor as blue collar jobs that are in decline. Meanwhile, sedentary jobs increasingly require computer and other cognitive skills that less educated older workers don’t have and that some workers lose as they age. And as job tenure has declined and fewer workers spend the bulk of their careers with one employer, older workers in poor health are at increased risk of losing their jobs.

Some of these explanations point to worsening employment prospects rather than worsening health. However, none suggest that financial incentives are causing more workers to voluntarily stop working or that examiners have relaxed the program’s stringent medical standards, though two musculoskeletal disorders were added to the list of conditions that can qualify as disabilities in 2002.

Moreover, viewing any increase in approvals for mental health and musculoskeletal conditions as problematic assumes that the diagnosis and treatment of these conditions were adequate in the past. This is clearly not true for mental health conditions and is doubtful even for musculoskeletal conditions causing debilitating pain. While there is also evidence of over-prescription and unmonitored use of drugs for mental health issues and pain driven by pharmaceutical firms seeking market expansion, there’s no evidence that patients with mild conditions are being approved for disability benefits.

Demographic trends, changes in reported diagnoses, growing health disparities, and declining employment prospects—whether these add up to a full explanation for the increased incidence of musculoskeletal disorders, they may not be answers to the most important question. Given the poor employment prospects of all disability applicants, including those who are denied benefits, instead of asking why more people with “subjective” conditions are applying for disability benefits rather than working, we should ask why more marginal applicants are applying for disability benefits rather than retiring or being supported by relatives.

This may seem like a subtle distinction, but it has important research and policy implications. Applying for disability benefits is time consuming, difficult, and may carry a stigma, so disabled workers avoid it if possible, especially when the outcome is uncertain—that is, for the roughly one in four applicants defined as “marginal” in Maestas, Mullen and Strand 2013. Having alternative sources of income, such as spousal earnings, pensions, and retirement savings, helps explain the Maestas et al. finding that high earners are less likely to apply for benefits unless their condition is severe enough to virtually guarantee being approved. Higher earners are also more likely to have desk jobs and work for employers willing to accommodate a disability.

Older workers with disabilities have fewer options than they did in the past because the share of private-sector workers with secure pensions has plummeted. Even among those with secure pensions, fewer have the option of retiring early or receive retiree health benefits before they are eligible for Medicare. Though private pensions, short-term disability insurance, and employer-provided health care can also tide workers over while they wait to be approved for disability benefits, it is likely that trends in employer-provided benefits have increased the share of older workers in poor health who apply for disability benefits rather than retiring early. Moreover, the share of Americans who get and stay married has declined, especially among lower-income workers, so fewer disabled workers may be supported by a spouse. Unmarried beneficiaries are much more reliant on SSDI benefits and are much more likely to be poor than married beneficiaries.

Other developments have had a more ambiguous effect on disability applications. One is the increasing importance of women’s earnings in dual-earner households, which may increase the number of women who apply for disability benefits and may decrease the number of men. Another is the passage of the Affordable Care Act, which guarantees affordable insurance for people with preexisting conditions. Health reform reduced the need to apply for SSDI benefits to become eligible for Medicare, but also made it easier for applicants to remain insured while they waited for SSDI and Medicare benefits.

Federal and state safety-net programs have also been cut or eliminated entirely. All told, there’s little doubt that disabled workers now have fewer alternatives to applying for SSDI. Undoubtedly, Duggan would frame these as increased “incentives” to apply, but there’s a difference between voluntarily choosing to stop working and applying for benefits to avoid destitution.

Duggan also assumes that the severity of mental health and musculoskeletal conditions is harder to assess and that beneficiaries with these conditions are more likely to be able to work. This is not as obvious as he suggests. It could be, for example, that workers with these conditions are reluctant to apply unless they have very severe impairments because they anticipate facing greater scrutiny.

If many applicants with musculoskeletal conditions have weak cases but hope to get lucky with a lenient examiner, we would expect to see a lower acceptance rate and a higher share of applicants who are marginal—that is, whose approval or rejection depends on the examiner. Maestas et al. actually find that applicants with mental health disorders and musculoskeletal conditions are about as likely to be accepted as applicants with other conditions. Applicants with musculoskeletal conditions are also less likely to be marginal, though those with mental health conditions are more likely to be marginal. Moreover, the very small negative effect on employment appears to be even smaller for beneficiaries with musculoskeletal disorders, though it is larger for those with mental health conditions.

Next: Disability and Employment Revisited

Previous: Does Disability Insurance Reduce Labor Force Participation?

In Senate testimony earlier this year, Stanford economist Mark Duggan claimed that Social Security Disability Insurance (SSDI) was an “important factor” in the decline in labor force participation in the United States relative to other industrialized countries. In an earlier blog post, I showed that even research cited by Duggan found that disability receipt had a negligible impact on overall employment. Is it still possible that SSDI has a noticeable impact on labor force participation, a measure that includes unemployed workers actively looking for work? It might, if you believe—as Duggan does—that the process of applying for benefits is enough to make people stop looking for work. But as will be detailed in this blog post, there are more likely explanations for the relative decline in labor force participation in the United States compared to Europe, including more supportive labor market policies in Europe.

In his testimony, Duggan points to an increase since 1990 in the prime-age (25-54) labor force participation rate for the EU-15 countries (countries that belonged to the European Union before it expanded into Eastern Europe) and a decline in the same measure in the United States. (Note that Duggan focuses on a measure that excludes older workers who have higher disability rates. The labor force participation of older workers is higher in the United States than in Europe, a fact that does not support Duggan’s claim that disability insurance is keeping Americans out of the labor force.)

Research that considers multiple causes for recent declines in participation in the United States generally finds that an aging workforce and unemployment—especially long-term unemployment—have been the main drivers (see, for example, research from the Council of Economic Advisors and Harvard University). The question is whether SSDI caused some unemployed workers to exit the labor force, or whether they would have exited regardless.

Looking at longer-term trends, the biggest difference between the United States and the EU-15 has been an increase in female employment and labor force participation in Europe, in which the adoption of family-friendly labor policies likely played a role. Prime-age employment rates in Europe are now similar to the United States (solid lines in the figures below) though they remain lower for older workers (dashed lines).

Figure 1

Figure 2

Though employment rates are similar, participation rates are now higher in Europe for prime-age workers. Almost all the difference (4.2 out of 4.7 percentage points in 2014) reflects higher unemployment in the EU-15 rather than higher employment. Because people receiving unemployment benefits are required to look for work, a contributing factor is likely to be better UI coverage in the EU-15 countries than in the United States, where many people who are out of work don’t qualify for unemployment benefits.

It might seem that any increase in disability rolls, even if due to an aging workforce, automatically reduces labor force participation, since disabled workers who aren’t working or actively looking for work are by definition outside the labor force. Even if true, this wouldn’t be sufficient reason to reform the program. But this assumes disabled beneficiaries aren’t working or actively looking for work (a few are) and—more importantly—that they would be working or actively looking if they weren’t receiving benefits (most wouldn’t be).

As discussed in a previous blog post, a closer look at research by economists Nicole Maestas and Kathleen Mullen of the RAND Corporation and Alexander Strand of the Social Security Administration suggests that 2.3 percent of new beneficiaries would be earning a living above the “substantial gainful activity” threshold two years later if they had been denied benefits. If these beneficiaries were working, would this have a measurable effect on prime-age labor force participation? It seems unlikely. For the sake of argument, consider the somewhat larger share (3.5 percent) of new beneficiaries who would earn any amount above $1,000 per year if they weren’t receiving benefits, and assume all would remain in the labor force. In addition, assume they live twice as long as the average beneficiary, making up 7 percent of the current beneficiary population. Multiplying 7 percent by the roughly 3 percent of the prime-age population who are receiving disability benefits, the potential increase in participation is roughly 0.2 percentage points, which is indistinguishable from statistical noise in labor force statistics.

Maestas et al. only consider marginal applicants who might be accepted or denied by different examiners. Disability benefits could also have a small effect on the employment and earnings of recipients who would have been accepted regardless of examiner, though the authors themselves describe any such effect as “statistically indistinguishable from zero.” Instead, Duggan suggests that the impact of SSDI on labor force participation may be larger than the small effect of benefit receipt on employment if the process of applying for benefits causes unemployed workers to drop out of the labor force whether or not they are ultimately approved for benefits.

Does it matter if people apply for disability benefits instead of remaining unemployed? (Either way, they’re out of work.) It might, if disability applicants are more likely to exit the workforce permanently and receive more generous government benefits. Though many unemployed workers also receive government benefits, disabled beneficiaries become eligible for Medicare two years after being approved for SSDI, which Duggan cites as a big financial incentive to apply for disability benefits.

In his testimony and elsewhere, Duggan claims SSDI “is to some extent serving as a form of long-term unemployment insurance for some workers.” However, economists Andreas Mueller of Columbia University, Jesse Rothstein of the University of California, Berkeley, and Till von Wachter of the University of California, Los Angeles compared the experiences of different states during the Great Recession and found no evidence that people applied for disability benefits when unemployment benefits ran out.

This doesn’t rule out the possibility that the SSDI application process is causing some unemployed workers to permanently exit the labor force, though there is no hard evidence of this. In a 2003 paper, Duggan and Massachusetts Institute of Technology economist David Autor estimated that SSDI caused a 0.5 percentage point decline in unemployment in the United States between 1984 and 1998. Over this period, unemployment trended downward in the United States while trending upward in Europe. As discussed in an earlier blog post, during the early part of this period, age-adjusted disability incidence also increased rapidly in the United States, rebounding after a Reagan-era retrenchment. But more recent trends don’t provide obvious support for the idea that SSDI has caused unemployed workers to exit the labor force. Prime-age unemployment increased more in the United States than in Europe in the last two recessions. Though unemployment in the United States declined faster after the Great Recession, disability incidence declined as well.

Other factors may obscure a causal link between SSDI and unemployment or labor force participation, which is tricky to prove or disprove because the direction of causality may go either way and because other factors, such as age composition and cyclical fluctuations in labor demand, affect both disability receipt and unemployment. In their 2003 paper, Autor and Duggan attempt to isolate the effect of SSDI on labor force participation using what’s known as an instrumental variable approach. They found that states with more low-wage workers in 1978—and therefore higher potential disability replacement rates due to Social Security’s progressive benefit formula—saw larger increases (or smaller declines) in disability rolls, and larger declines (or smaller increases) in labor force participation for high-school dropouts between 1978 and 1998  than states with lower replacement rates.

However, the technique used by Autor and Duggan relied on the assumption that potential replacement rates in 1978 measured workers’ incentive to apply for disability benefits without being correlated with other factors that affected disability and labor force participation trends. Though Autor and Duggan focus on changes within states to account for unobserved differences between them, it’s likely that states with more low-wage workers in 1978 were also more affected by other factors that contributed to worsening health and declining job prospects for high-school dropouts, including growing health disparities, rising educational attainment, and an aging population (Autor and Duggan only partly control for age).

With the benefit of hindsight, what is striking about Autor and Duggan’s 2003 paper is how many factors they didn’t consider, let alone rule out, including the rising share of women insured for benefits. In a later (2006) paper, Autor and Duggan do consider the share of women who are insured, but make no attempt to isolate the causal effect of SSDI receipt on unemployment or participation. Thus, it’s not clear whether Duggan has any basis for the claim in his Senate testimony that SSDI was an “important factor” in the decline in labor force participation in the United States compared to other industrialized countries.

In general, the only obvious conclusion to be drawn from comparing the United States to other countries is that disability benefits in Western Europe and most other advanced economies remain more generous, despite recent cutbacks. Around 6 percent of the working-age population in the United States receives disability benefits, the same as the OECD average. However, the United States spends less on benefits: 0.8 percent of GDP, compared to an OECD average of 1.1 percent. The major European economies (EU-15) spend even more (1.3 percent of GDP) (author’s analysis of OECD data).  These figures don’t include other benefits received by disability beneficiaries, but a more inclusive measure is not likely to change the overall picture. As a result, the United States has the dubious distinction of having the highest disability poverty rate in the OECD. Though this partly reflects a higher overall poverty rate, the poverty rate for disabled Americans is high even relative to overall poverty in the United States.

Next: Do Disability Trends Reflect a Liberalization of the Program’s Medical Criteria?