REPORT Affordable Care ActInequalityUnited StatesWorkers

Toward Black Full Employment: A Subsidized Employment Proposal


Enough Is Enough: Ending 60 Years of High Black Joblessness

In 1963, Dr. Martin Luther King Jr., gave his famous “I Have a Dream” speech at the March on Washington for Jobs and Freedom. As the title of the march indicates, jobs — or rather, the high rate of joblessness for Black people — was one of the motivating factors behind the march. Next year will be its 60th anniversary, and yet the problem of high Black joblessness remains. In fact, the problem has only worsened since 1963. The Center for Economic and Policy Research (CEPR) will be issuing a series of reports highlighting the problem of Black joblessness and proposing solutions. This is the first report in the series. CEPR calls on policymakers to enact policies to address this long-standing problem.

From at least as early as the 1960s to today, the Black unemployment rate has been about twice the white unemployment rate.1 This stable unemployment rate disparity means that no intervention to achieve equal employment opportunities for Black Americans has had any significant success in the past 60 years.

A large subsidized employment program could finally break this two-to-one, Black-to-white unemployment rate ratio. Subsidized employment programs use government funding to cover some or all of the wage costs for hiring employees. By substantially reducing the cost of employees to organizations, this policy increases the demand for workers. A subsidized employment program targeting job creation in communities suffering from persistently high rates of joblessness would improve employment prospects for everyone living in such communities, but it would disproportionately benefit Black people because Black communities tend to have high rates of joblessness.

In recent years, a number of economists and economic policy organizations have put forth subsidized employment proposals.2 This report will discuss the potential of a federal subsidized employment program to reduce the high joblessness among Black Americans and sketch some design features that would make that program more effective at addressing Black joblessness.

Key Findings

  • The white-Black jobs gap is one of the root sources of Black American economic hardship. Currently, Black America has a jobs deficit of nearly 1.5 million jobs. This jobs deficit translates to roughly $70 billion in income lost to the community each year.
  • In 1939, the subsidized employment program by the Works Progress Administration (WPA) employed over 400,000 Black Americans. The US Black population is much larger today than it was in 1939, so an equivalent number of jobs today would be about 1.4 million.
  • Subsidized employment programs have a positive track record of putting the jobless back to work. A recent meta-analysis concludes that “[w]age subsidies show the greatest impact on labor earnings and employment” among worker-centered employment programs.
  • A subsidized employment program can move the Black community toward full employment. To best serve this group, the program should:
    • target communities with persistently low prime-age employment;
    • use federal funding and administration;
    • provide 10-year subsidies with economic development support;
    • provide technical assistance to the most economically disadvantaged communities;
    • only subsidize net new jobs at prevailing wages;
    • make public, nonprofit, and for-profit organizations eligible for subsidies;
    • ensure targeted hiring;
    • give hiring preference to the long-term jobless; and
    • ensure that all types of workers are eligible.
  • A subsidized employment program will be most effective when the national unemployment rate is low due to other complementary job-creation policies, when there are skills development and wraparound services for the targeted communities, and when there are strong anti-discrimination policies in the labor market. Though these policies are useful and important, they are not a substitute for a targeted subsidized employment program.

The Need for Subsidized Employment for Black Americans

As bad as the problem of Black joblessness looks when examining the unemployment rate, the problem is actually far worse. The unemployment rate undercounts Black joblessness. The technicalities of the official definition of unemployment means that individuals can lack a job, be willing to work, and yet still not be counted as unemployed.3 Groups experiencing persistent challenges finding jobs are most likely to be undercounted by the unemployment rate. For these groups, the employment-to-population ratio, which is also called the employment rate, provides a more accurate picture of joblessness.

Figure 1 compares the white-Black employment rate and unemployment rate jobs gaps by sex and age category for the average of the first seven months of 2022. For all the Black subgroups except one, the employment rate gap is substantially larger than the unemployment rate gap. The largest in terms of the number of jobs is for prime-age (25 to 54 years old) Black men. At 736,000 jobs, their employment rate gap is 543,000 jobs greater than the unemployment rate gap. The employment rate gap for Black women who are 16 to 24 years old is 311,000 jobs — or 211,000 jobs over the unemployment rate gap. For Black men who are also 16 to 24 years old, it is 246,000 jobs — or 143,000 jobs over the unemployment rate gap.

Figure 1

The exception is for prime-age Black women. When the overall labor market is tight, Black women and white women have similar employment rates.4 For the first seven months of 2022, the average employment rate for prime-age Black women was slightly higher than for prime-age white women, leading to a negative employment rate gap. Although proportionally more prime-age Black women are working than prime-age white women, prime-age Black women still have a higher unemployment rate. This means that there are still proportionally more prime-age Black women actively looking for work but unable to find it than prime-age white women. For prime-age Black women then, the unemployment rate is appropriate for assessing the jobs gap.

This high level of joblessness causes tremendous hardship for Black communities. Summing the employment rate jobs gaps for women and men 16 to 24, men 25 to 54, and the unemployment rate jobs gap for prime-age Black women yields a total of 1,477,000 jobs. This is the size of the total white-Black jobs gap during this current period with exceptionally low unemployment nationally. It would be larger with a weaker labor market. This total jobs gap value is 2.5 times the comparable number just based on the unemployment rate. Today, a rough estimate of the income lost from this joblessness is $70 billion.5 If Black America were to obtain an additional $70 billion each year, there would be considerably less poverty and fewer problems stemming from poverty in this population. The white-Black jobs gap is one of the root sources of economic hardship for Black Americans.

Subsidized employment programs have a positive track record of putting the jobless back to work. A recent meta-analysis of 102 randomized control trials for four categories of worker-centered employment programs concludes that “[w]age subsidies show the greatest impact on labor earnings and employment.” 6

Examples from the past demonstrate that a subsidized employment program can be designed to help achieve Black full employment today.7 For example, the Temporary Assistance for Needy Families (TANF) Emergency Fund was used to create over a quarter of a million jobs.8 The Comprehensive Employment and Training Act (CETA) created more than three quarters of a million jobs.9 Since many poor Black communities can fairly be described as economically depressed, the Great Depression-era examples may be best. In 1939, the subsidized employment program by the WPA employed over 400,000 Black Americans.10 The US Black population is much larger today than it was in 1939, so an equivalent number of jobs today would be about 1.4 million.11 This calculation does not include the nearly 200,000 Black youth who found subsidized employment in the Civilian Conservation Corps (CCC) between 1933 and 1942.12

Key Features of a Subsidized Employment Program for Black Americans

Target Communities with Persistently Low Prime-Age Employment

A subsidized employment program should target communities with persistently low prime-age employment — regardless of race. Among all racial groups, there are communities that suffer from high rates of joblessness. Nationally, Native Americans have an even higher rate of joblessness than Black Americans.13 Joblessness is also a serious problem in the majority white states of West Virginia and Kentucky.14 Economist Timothy J. Bartik and others have shown that there are “severely distressed” labor markets in every state in the nation.15 Bartik also finds that 7.3 percent of the white population, 11.8 percent of the Hispanic population, and 21.7 percent of the Black population reside in low-employment neighborhoods.16 All racial groups will benefit from a program targeted to low employment areas, but the benefit will be disproportionate to groups suffering from higher rates of joblessness.

Figure 2

For a subsidized employment program to be most effective at addressing Black joblessness, it must be able to target a sufficiently small geographic area. For some areas, a large geographic area may be appropriate, but there must be enough flexibility that targeting would be able to reach the segregated, majority Black areas in many of our major cities. The rationale for this guidance is illustrated in Figure 2. The figure shows that, for the 2015 to 2019 period, the Washington, DC, metropolitan area had an employment rate 5.4 percentage points above the national average. The city of Washington, DC, had a rate 3.6 percentage points above the national average. Both of these areas could be considered high-employment areas and would be excluded from a subsidy program if the metropolitan area or city were the target geography. Wards 7 and 8 — within the city of Washington, DC — however, were 11.2 percentage points below the national average (and 16.6 percentage points below the metropolitan-area average). Wards 7 and 8 are low-employment areas, and they are about 90 percent Black.17 Only if smaller geographic areas were allowed for targeting would areas such as Wards 7 and 8 receive subsidies. If a subsidized employment program is designed to target joblessness based on city or metropolitan-area employment rates (or even larger geographies such as commuting zones),18 substantial amounts of Black joblessness will be missed.

Employment Equity Could Mean an Additional $550 Million for the Majority Black Wards 7 and 8 in Washington, DC

As shown in Figure 2, for the 2015 to 2019 period, the prime-age employment rate for Wards 7 and 8 in Washington, DC, was 14.8 percentage points below the city average. If that gap could be closed, those wards would have roughly an additional $550 million a year.19

As already mentioned, these wards are about 90 percent Black. Two-thirds of the children in Wards 7 and 8 live in households with very low or low incomes (below 200 percent of the official poverty line).20 With the additional income from closing the employment rate gap, many fewer Black children would live in poverty in these wards.


$126 Million a Year to Combat White Rural Poverty in Carter County, Kentucky

With the decline of coal, eastern Kentucky has become a place of persistent rural poverty.21 Although Kentucky has a relatively low employment rate (for 16-years-old and over, 55.7 percent), Carter County’s employment rate was 15.7 percentage points lower than the state’s average for the 2015 to 2019 period.22 The county is nearly all white, and 18.5 percent of residents live below the official poverty line.23 Closing that employment rate gap could add $126 million a year to the county,24 which would bring many positive changes.


Closing the Jobs Gap in Imperial County, California, Could Generate an Additional $536 Million for Local Latino Workers

The prime-age employment rate in Imperial County, California, in the 2015 to 2019 time period was 60.8 percent, 16.4 percentage points below the state average (77.2 percent).25 Employment in Imperial County reaching parity with the state would generate approximately $536 million per year in additional income.26

Imperial County is roughly 86 percent Latino and over 18 percent of residents live below the official poverty line.27 Closing the jobs gap between Imperial County and the state of California through a subsidized employment program could provide a pathway out of poverty for many Latino households living in this area.


Subsidized Employment in Apache County, Arizona, Could Generate $598 Million for Native American Families

Poverty rates are notoriously high in Indigenous communities throughout the United States. Research suggests that this poverty is driven largely by joblessness.28 Apache County, Arizona, which encompasses part of the Navajo Nation and Fort Apache Indian Reservation, is 74.5 percent American Indian.29 During the 2015 to 2019 time period, just 33.9 percent of its residents (ages 16 and over) were employed. Employment in the state of Arizona for the same time period sat at 55.9 percent, 22 percentage points higher than Apache County.30

A subsidized employment program that closed this employment gap would generate close to $600 million in additional income for workers in Apache County,31 where nearly one third of residents live below the official poverty line.32


Use Federal Funding and Administration

Federal, state, or local government could fund a subsidized employment program, but federal funding and administration may be preferable to better serve Black joblessness. As occurred with Medicaid expansion under the Affordable Care Act (ACA), several conservative-led states chose not to expand Medicaid, causing disproportionate harm to Black Americans.33 More recently, Mississippi has ended its participation in a federal rental assistance program — although it still has access to $130 million in federal funding, and there is still a need.34 Even with the Great-Depression-era Works Progress Administration projects, there was sometimes conflict between the federal and state governments. Though WPA funding was supposed to be used for employment, some state governments only wanted to provide cash assistance without establishing work projects. This caused the federal government to assume more complete control of the administration of WPA projects and bypass state governments to work directly with municipalities.35 The benefits of a subsidized employment program should not be allowed to be blocked by state governments.

In the event that there is no federal program, state and local governments should enact subsidized employment programs on their own. Even if there is a federal program, state and local governments might choose to add to the available federal funding since the programs bring considerable social and economic benefits. Bartik argues that “when we consider how job creation and higher employment rates affect a community — lower substance abuse, lower crime, stronger families and child development, better job skills, better fiscal conditions and public services, and so forth — the required social benefit measure might be twice the wage bill in a distressed community.”36

Provide 10-Year Subsidies with Economic Development Support

A federal subsidized employment program for persistently low-employment communities should provide community funding for at least a 10-year period.37 The goal is to use this funding along with other economic development activities to try to revitalize economically depressed communities. Very short-term funding cannot do this.

Individuals should be able to work in a subsidized job for at least a year. Ideally, the job would continue after the expiration of the subsidy. As the Georgetown Center on Poverty and Inequality reports, “the rigorously evaluated programs . . . that have longer subsidies have the most consistent record of improving employment and earnings.”38 It appears that longer subsidies yield longer-term benefits for workers.

While the vast majority of the program’s funds should directly support employment, it would be beneficial for a small portion to also be used for broad economic planning and development. It is important to have a strategy to sustain employment beyond subsidies.

Provide Technical Assistance to the Most Economically Disadvantaged Communities

Competitive grants can replicate existing socioeconomic inequality, and they should therefore not be used for allocating funding for a subsidy program. With competitive grants, communities with greater socioeconomic resources tend to also have better resources for securing grants. Since the goal is to lift up the neediest communities, competitive grants would undermine the process by providing an advantage to the better-off communities.

The subsidized employment program should work to recruit and provide technical assistance to the most disadvantaged communities. The neediest communities may not have the resources to create a successful public-sector job program and community development plan on their own. Resources should be made available to these communities to use if they wish to do so.

Only Subsidize Net New Jobs at Prevailing Wages

The subsidized employment program must only support net new jobs; it must not replace or displace existing jobs or workers. For example, this means that an organization that has 10 employees can only receive a subsidy for new employees over 10. The wage rate should be based on the prevailing wage.

Make Public, Nonprofit, and For-Profit Organizations Eligible for Subsidies

The subsidized employment program should provide wage subsidies to public, nonprofit, and for-profit organizations within the target area and in nearby areas. If the target area is low in terms of the number of local businesses relative to other comparable nearby areas, the subsidies should be used to encourage economic development.

Communities should be encouraged to think creatively about expanding their public sector. Communities suffering from high rates of joblessness typically lack services and amenities found in wealthier communities. The subsidized employment program should be used to address this issue. For example, many urban communities could benefit from expanding and investing in their parks department. The American Society of Civil Engineers gives America’s parks a D-plus grade,39 so in many communities, there will be considerable room for improvement. There are reasons why investing in parks and other types of green spaces are particularly important to Black Americans:

Because the Black population is disproportionately urban, public parks can be especially valuable to this population. Parks can be the only green spaces in an urban community — and therefore the only place to receive the mental health benefits of green spaces. Urban areas can become heat islands, and because of climate change, we can expect this dynamic to worsen. Trees and other vegetation found in parks can significantly cool areas. Parks that include rain gardens, floodplains, and estuaries can help mitigate the effects of flooding, which is expected to worsen with climate change. Converting empty lots and abandoned, decrepit buildings into green spaces have been found to reduce crime.40

This is just one area where a subsidized employment program can be used to support the local public sector to do work for the public good.

Some subsidized employment proposals have only called for public sector jobs.41 However, there is already a positively received history of wage subsidies for jobs in the private sector.42 Also, the government regularly provides subsidies, usually in the form of tax incentives, to the private sector with the hope that significant job creation will follow. Subsidized employment programs are a more efficient and reliable way for the government to obtain private sector job creation. Allowing more types of organizations to receive subsidies will likely make it easier to have a large jobs program. Also, private sector jobs may be more likely to remain after the expiration of jobs subsidies.

It may be appropriate to place restrictions on the type of private organization that can receive a subsidy. Subsidies could be limited to smaller businesses. The economic multiplier in the local community for small business subsidies tends to be larger than for large businesses.43

Private sector subsidies could be used to spur economic development. For example, there is concern that many poor, predominantly Black communities are “food deserts” without access to affordable, healthy food. If an entrepreneur is told that an area that is a food desert will be experiencing increased household incomes because it is the target for a large-scale job subsidy program, and that a new grocery store would be eligible for wage subsidies, then that entrepreneur may be convinced that it is profitable to open a grocery store in the area. The time-limited subsidy may spur needed business development in the target area. Essentially, this is similar to other types of business development grants and tax incentives, but there are guarantees that community members receive benefits with this design.

Ensure Targeted Hiring

It is possible for the subsidized jobs created in a low-employment community to be filled by residents of a neighboring higher-employment community or by workers recruited from afar by employers. To prevent this scenario, employers should be required to obtain employees from a local community-based referral organization(s) in order to obtain a wage subsidy. Under community benefits agreements, this type of organization is referred to as the “first source” office because this organization is the first place businesses must look for employees.44

This “first-source” type office could be part of a high school or community college or an organization providing job training or a church or other community organization. The office should actively recruit jobless community members to be job candidates. It can assist in matching job candidates to job openings. A federal subsidized employment program should include funding for “first-source” type offices and include other mechanisms to ensure that individuals living in the targeted community benefit from the new jobs created.

Give Hiring Preference to the Long-Term Jobless

A subsidized employment program should give preference to the long-term jobless to bring them back into the active labor market. This will require active recruitment from a “first-source” type office or other entity. The more recently unemployed should also have a preference over individuals who are currently employed. Individuals currently employed in bad jobs, however, should not be excluded from opportunities to transition to a better subsidized job.

Ensure That All Types of Workers Are Eligible 

It is important to recognize that there is a large number of Black Americans who are jobless. Subsidized jobs programs should be designed for a diverse Black population. Black jobless adults may be made up disproportionately of high school dropouts, but there are likely high school and college graduates too. There are individuals who have been formerly incarcerated as well as individuals who have not. There are individuals with disabilities as well as those without. Ideally, within a specific geography, an assessment would be made of the demographics of the jobless so that the subsidy program can include the appropriate training and support services that are needed.

Inflation and Other Risks of a Federal Job Guarantee

A permanent federal job guarantee for all is a powerful and popular idea.45 If everyone was guaranteed a good job at all times, this would indeed produce full employment for Black Americans. Because it is a universal guarantee, it would be a much larger, more costly program than a targeted subsidized employment program. In addition to a greater cost, it would also likely have greater political challenges and economic risks.

A federal job guarantee runs the risk of being inflationary.46 The much-feared wage-price spiral that produces continually increasing inflation probably would not happen,47 but it could still lead to substantial price increases. Tight labor markets and the wage increases they induce are good, but a federal job guarantee runs the risk of making already tight labor markets too tight. To use the example in Figure 2, a federal job guarantee would exert pressure to increase employment in the areas of the Washington, DC, metropolitan area that already have above-average rates of employment. This would spur wage increases, and those wage increases could potentially lead to rising prices.

In contrast, a jobs subsidy targeted to Wards 7 and 8 would not put upward wage pressures on the metropolitan area since there is a high rate of joblessness in those wards and because the overall increase in employment and income is small. If one were to increase the prime-age employment rate in Wards 7 and 8 to match the city rate, that would require fewer than 10,000 jobs. This would represent a bit less than a 3-percent increase in the city’s employment, and, ideally, many of the newly employed would have not been actively participating in the labor market prior to the program. For the metropolitan area, it would be about a 0.3-percent increase in employment. The percent increase in income would be significantly less than the increase in employment because these will be relatively low-wage workers. Thus, it would not be much more difficult for employers to find workers after the employment program than before the program, and the slight increase in demand for goods locally should not lead to any scarcity.

On a national level, the effects of a subsidized employment program would be similar to its effect in Washington, DC. If the program is successful, it could increase the number of employed by 2 percent and the national income by 1 percent.48 Also, these increases would likely take time to occur, which would also mitigate their effects. Thus, a subsidized employment program is not likely to increase inflation due to higher labor costs or to a large increase in the demand for goods and services.

Because a federal job guarantee is a universal program, its size and economic effect would be significantly larger than for a targeted program. It could be five times the effect of a targeted program49 — or even greater. Some job guarantee proposals call for a wage greater than what is currently available to a third of the US workforce.50 If these proposals were implemented too rapidly, they would cause tremendous economic disruption since everyone in lower-paid jobs — a third of the workforce — would leave their jobs for the higher compensation under the federal program. For a job guarantee to not cause disruption, it would need to be gradually phased in. One way to do this could be with a targeted subsidized employment program.

It is worth noting that there are no functioning models of a job guarantee in Western capitalist countries. Even among the countries that are much better than the United States at providing for the basic needs of their citizens and at reducing poverty and inequality, there are not yet any job guarantees.51 A subsidized employment program could be a way to experiment with the job guarantee idea without fully committing to it.

Skills Development and Wraparound Services Are Necessary, But Not Sufficient

Figure 3

Economic development planning and investments, job training programs, and wraparound services and supports (e.g., childcare, transportation, substance abuse treatment) should be incorporated into subsidized employment plans, but they are not a substitute for an employment program. Only subsidized employment programs guarantee jobs. Job training — unless it is via a recognized apprenticeship program — is not a job.

It is important to be clear that the high rate of joblessness among Black Americans is not simply the result of a lack of skills or a spatial mismatch.52 Because of this fact, increasing the skills of Black Americans or providing transportation and other wraparound services are not likely to be enough to produce a large increase in employment.

Figure 3 provides support for this view. The data are for youth, 16 to 24 years old, who are no longer enrolled in school. Although both white and Black youth who are high school dropouts have little work experience and little formal education, the Black youth dropouts are about twice as likely to be unemployed as the white youth dropouts. This difference in unemployment rates is not likely to be due a skill differential. This same Black-to-white unemployment ratio disparity is also present at higher levels of educational attainment. Equal educational attainment does not lead to equal employment outcomes for Black Americans.

Sectoral training programs are among the better job training programs, but they are mixed in their ability to produce benefits for Black participants that are comparable to non-Black participants.53 One evaluation illustrates the sometimes poor outcomes for Black participants in these programs. The Per Scholas sectoral training program provides training to become a computer technician to residents of the South Bronx in New York City. An evaluation of the Per Scholas training program found that the Black people in the program received an A+ certification at rates comparable to other groups. The Black American A+ certification rate was 53 percent. The Latino rate was 54 percent, and the formerly incarcerated rate was 50 percent. In spite of this similarity in A+ certifications, the Black group did not have statistically significant improvements in their employment rates after the program relative to their control group while the Latino and formerly incarcerated groups did.54 For Black Americans, having the right skills may not be enough to obtain a job.

Since the Per Scholas program is based in one geographic area, a spatial mismatch explanation for the worse outcome for Black people is not likely. More careful and rigorous research has also concluded that “space alone plays a relatively minor role in low black male employment rates.”55

Audit studies in which employers are presented with Black and white candidates with the same qualifications consistently show that employers have a significant white preference in hiring.56 A subsidized employment program targeted to segregated, predominantly Black communities can help to provide equal employment opportunity.

Complementary Policies

A subsidized employment program for moving Black Americans toward full employment will be most effective when complemented with other policies that work to produce low unemployment nationally and with a strong commitment to anti-discrimination policies in hiring. It is easier to achieve Black full employment during periods when the overall unemployment rate is low.57 A prior analysis of the white-Black jobs gap for prime-age men showed that during a low-unemployment period, the jobs gap was about three-quarters of a million jobs but was a million jobs during a high-unemployment period.58 The smaller the job gap, the easier it is to fill.

There are a number of policies that would help to keep the national unemployment rate low. Among them are encouraging the Federal Reserve not to increase interest rates prematurely and unnecessarily; investing in infrastructure and climate change mitigation, which will create many jobs; reducing the US trade deficit; reducing work hours by having stronger vacation, paid leave, and sick leave policies; and encouraging work-sharing during recessions.59 There are likely other helpful policies.

As the prior section suggested, anti-Black racial discrimination is a factor behind the high rate of joblessness for Black Americans. As the persistence of the two-to-one, Black-to-white unemployment rate ratio indicates, we have done little to reduce this. Thirty years of rigorous employment audit studies show continuing anti-Black racial discrimination in the labor market.60 Strong anti-discrimination policies are necessary to ensure that jobs created are fairly accessible to all.


High levels of joblessness have plagued Black communities for the past two generations. This situation has caused these communities to suffer from high rates of poverty and from the many problems that stem from concentrated poverty. A large, federally funded subsidized employment program could finally break this pattern and move these communities toward full employment.


The author thanks Kendra Bozarth, Karen Conner, Lisa Burnam, Sarah Rawlins, and Matt Sedlar for editorial assistance, and Eileen Appelbaum for editorial review. This research was funded by the Annie E. Casey Foundation. We thank them for their support but acknowledge that the findings and conclusions presented in this report are those of the author alone, and do not necessarily reflect the opinions of the Foundation.


Altman, Daniel. “The Myth Of The Wage-Price Spiral.” Forbes, July 11, 2022. https://www.forbes.com/sites/forbesbusinesscouncil/2022/07/11/the-myth-of-the-wage-price-spiral/.

American Society of Civil Engineers. 2021. “Report Card for America’s Infrastructure: A Comprehensive Assessment of America’s Infrastructure.” https://infrastructurereportcard.org/wp-content/uploads/2020/12/National_IRC_2021-report.pdf

Argitis, Giorgos, and Nasos Koratzanis. “A European Job Guarantee to Foster Wellbeing.” Social Europe, January 7, 2021. https://socialeurope.eu/a-european-job-guarantee-to-foster-wellbeing.

Austin, Algernon. 2013. “The Unfinished March: An Overview.” Economic Policy Institute. https://www.epi.org/publication/unfinished-march-overview/.

________. 2021a. “More Than Roads and Bridges: A Comprehensive Approach to Sound Infrastructure Is an Important Counter to Historic Racial Inequity.” Thurgood Marshall Institute. https://tminstituteldf.org/wp-content/uploads/2021/07/LDF_07232021_InfrastructureTMIBrief-8.pdf

________. 2021b. “The Jobs Crisis for Black Men Is a Lot Worse Than You Think.” Center for Economic and Policy Research. https://cepr.net/report/the-jobs-crisis-for-black-men-is-a-lot-worse-than-you-think/.

________. 2022. “High Joblessness for Black Youth: More Than 500,000 Jobs Are Needed.” Center for Economic and Policy Research. https://cepr.net/report/high-joblessness-for-black-youth-more-than-500000-jobs-are-needed/.

Austin, Benjamin, Edward Glaeser, and Lawrence Summers. 2018. “Jobs for the Heartland: Place-Based Policies in 21st-Century America.” Brookings. https://www.brookings.edu/wp-content/uploads/2018/03/AustinEtAl_Text.pdf.

Baker, Dean, and Jared Bernstein. Getting Back to Full Employment: A Better Bargain for Working People. Washington, DC: Center for Economic and Policy Research, 2013.

Bartik, Timothy J. 2022a. “How State Governments Can Target Job Opportunities to Distressed Places.” W.E.Upjohn Institute for Employment Research.  https://research.upjohn.org/cgi/viewcontent.cgi?article=1044&context=up_policybriefs.

________. 2022b. Testimony to the Select Committee on Economic Disparity and Fairness in Growth Hearing on “Bringing Prosperity to Left-Behind Communities: Using Targeted Place-Based Development to Expand Economic Opportunity,” May 11.  https://fairgrowth.house.gov/sites/democrats.fairgrowth.house.gov/files/documents/Bartik%20HHRG-117-EF00-Wstate-BartikT-20220511.pdf.

Bashay, Molly, and Asha Banerjee. 2020. “Building an Equitable Recovery for Workers and Families: CLASP Principles for Subsidized Employment.” Center for Law and Social Policy.   https://www.clasp.org/wp-content/uploads/2022/04/202020CLASP20Principles20for20Subsidized20Employment.pdf.

Bivens, Josh. 2018. “Recommendations for Creating Jobs and Economic Security in the US: Making Sense of Debates about Full Employment, Public Investment, and Public Job Creation.” Economic Policy Institute. https://www.epi.org/publication/creating-jobs-and-economic-security/.

Camardelle, Alex, Harin Contractor, Paul Elam Jr., Colleen Graber, and Spencer Overton. 2022. “Improving Training Evaluation Data to Brighten the Future of Black Workers.” Joint Center for Political and Economic Studies. https://jointcenter.org/improving-training-evaluation-data-to-brighten-the-future-of-black-workers/.

Center for American Progress. 2018. “Blueprint for the 21st Century: A Plan for Better Jobs and Stronger Communities.” https://www.americanprogress.org/article/blueprint-21st-century/.

Cole Jr., Olen. The African-American Experience in the Civilian Conservation Corps. Gainesville: University Press of Florida, 1999.

Dutta-Gupta, Indivar, Kali Grant, Julie Kerksick, Dan Bloom, and Ajay Chaudry. 2018. “Working to Reduce Poverty: A National Subsidized Employment Proposal.” RSF: The Russell Sage Foundation Journal of the Social Sciences February 2018, 4 (3) 64-83. DOI: https://doi.org/10.7758/RSF.2018.4.3.04.

Dutta-Gupta, Indivar, Kali Grant, Matthew Eckel, and Peter Edelman. 2016. “Lessons from 40 Years of Subsidized Employment Programs: A Framework, Review of Models, and Recommendations for Helping Disadvantaged Workers.” Georgetown Center on Poverty and Inequality.   https://www.georgetownpoverty.org/issues/employment/lessons-learned-from-40-years-of-subsidized-employment-programs/.

Eckel, Matthew, Jess Belledonne, Bre Bambrick, Kali Grant, Zachariah Oquenda, and Indi Dutta-Gupta. 2021. Subsidized Employment Programs [Workbook]. Georgetown Center on Poverty and Inequality. https://www.georgetownpoverty.org/issues/subsidized-employment-programs/.

Escudero, Verónica, and Elva López Mourelo. 2017. “The European Youth Guarantee: A Systematic Review of Its Implementation across Countries,” Research Department Working Paper No. 21. International Labour Office. https://www.ilo.org/global/research/publications/working-papers/WCMS_572465/lang–en/index.htm.

Gross, Julian, Greg LeRoy, and Madeline Janis-Aparico. 2005. “Community Benefits Agreements: Making Development Projects Accountable.” Good Jobs First. https://goodjobsfirst.org/community-benefits-agreements-making-development-projects-accountable/.

Harker, Laura. 2021. “Closing the Coverage Gap a Critical Step for Advancing Health and Economic Justice.” Center on Budget and Policy Priorities. https://www.cbpp.org/research/health/closing-the-coverage-gap-a-critical-step-for-advancing-health-and-economic-justice.

Harvey, Philip. 2011. “Back to Work: A Public Jobs Program for Economic Recovery.” Dēmos.  https://www.demos.org/sites/default/files/publications/Back_To_Work_Demos.pdf.

Hellerstein, Judith K., David Neumark, and Melissa McInerney. 2008. “Spatial Mismatch or Racial Mismatch?” Journal of Urban Economics 64(2), September: 464-479. https://www.sciencedirect.com/science/article/abs/pii/S0094119008000387

Henderson, Kaitlyn. 2022. “The Crisis of Low Wages in the US: Who Makes Less Than $15 an Hour in 2022?” Oxfam. https://webassets.oxfamamerica.org/media/documents/low_wage_report_2022_final.pdf

Horowitz, Julia. “Job Guarantees and Free Money: ‘Utopian’ Ideas Tested in Europe as the Pandemic Gives Governments a New Role.” CNN Business, November 23, 2022. https://www.cnn.com/2020/11/23/economy/universal-basic-income-europe-pandemic/index.html.

Hendra, Richard, David H. Greenberg, Gayle Hamilton, Ari Oppenheim, Alexandra Pennington, Kelsey Schaberg, and Betsy L. Tessler. 2016. “Encouraging Evidence on a Sector-Focused Advancement Strategy: Two-Year Impacts from the WorkAdvance Demonstration.” MDRC. https://www.mdrc.org/publication/encouraging-evidence-sector-focused-advancement-strategy.

Institute for Policy Research. 2020. “What Drives Native American Poverty? Sociologist Beth Redbird’s Research Points to Job Loss, Not Education, as a Key Driver.” Northwestern University. https://www.ipr.northwestern.edu/news/2020/redbird-what-drives-native-american-poverty.html.

Levy Yeyati, Eduardo, Martin Montane, and Luca Sartorio. 2019. “What Works for Active Labor Policies?” CID Working Paper Series 2019.358. Harvard University, Cambridge, MA, July. https://nrs.harvard.edu/URN-3:HUL.INSTREPOS:37366396.

Lowery, Annie. “What’s the Matter With Eastern Kentucky?” New York Times Magazine, June 26, 2014. https://www.nytimes.com/2014/06/29/magazine/whats-the-matter-with-eastern-kentucky.html.

McCausland, Phil.“Mississippi Will Send Back Fed’s Rental Aid, Even as Housing Needs Remain High.” NBC News, August 13, 2022. https://www.nbcnews.com/news/us-news/mississippi-will-send-back-cash-federal-rental-aid-program-even-renter-rcna42547.

Maguire, Sheila, Joshua Freely, Carol Clymer, Maureen Conway, and Deena Schwartz. 2010. “Tuning In to Local Labor Markets: Findings from The Sectoral Employment Impact Study.” Public/Private Ventures. https://ppv.issuelab.org/resource/tuning-in-to-local-labor-markets-findings-from-the-sectoral-employment-impact-study.html.

Paul, Mark, William Darity, Jr., and Darrick Hamilton. 2018. “The Federal Job Guarantee: A Policy to Achieve Permanent Full Employment.” Center on Budget and Policy Priorities. https://www.cbpp.org/sites/default/files/atoms/files/3-9-18fe.pdf.

Pavetti, Ladonna, Liz Schott, and Elizabeth Lower-Basch. 2011. “Creating Subsidized Employment Opportunities for Low-Income Parents: The Legacy of the TANF Emergency Fund.” Center on Budget and Policy Priorities. https://www.cbpp.org/research/creating-subsidized-employment-opportunities-for-low-income-parents.

Porter, Eduardo. “Who Discriminates in Hiring? A New Study Can Tell” The New York Times, January 29, 2021. https://www.nytimes.com/2021/07/29/business/economy/hiring-racial-discrimination.html.

Quillian, Lincoln, Devah Pager, Ole Hexela, and Arnfinn H. Midtbøen. 2017. “Meta-Analysis of Field Experiments Shows No Change in Racial Discrimination in Hiring Over Time.” PNAS, October 10, vol. 114, no. 41. www.pnas.org/cgi/doi/10.1073/pnas.1706255114.

Ruggles, Steven, Sarah Flood, Ronald Goeken, Megan Schouweiler, and Matthew Sobek. 2022. IPUMS USA: Version 12.0 [dataset]. Minneapolis, MN: IPUMS. https://doi.org/10.18128/D010.V12.0.

________. “Majority of Voters Support a Federal Jobs Guarantee Program.” The Hill, October 30, 2019. https://thehill.com/hilltv/468236-majority-of-voters-support-a-federal-jobs-guarantee-program/.

US Bureau of Labor Statistics. 2015. “Frequently Asked Questions.” https://www.bls.gov/cps/faq.htm#Ques5.

US Census Bureau. 1943. “Population: Characteristics of the Nonwhite Population by Race.” Sixteenth Census of the United States: 1940. US Department of Commerce. https://www2.census.gov/library/publications/decennial/1940/population-nonwhite/population-nonwhite.pdf.

US Bureau of the Census. 1940. Population Characteristics of The Nonwhite Population By Race. Sixteenth Census of the United States. https://www2.census.gov/library/publications/decennial/1940/population-nonwhite/population-nonwhite.pdf.

________. 2021. “Table P-8. Age — Black People, by Median Income and Sex: 1967 to 2020,” Historical Income Tables: People. US Department of Commerce. https://www.census.gov/data/tables/time-series/demo/income-poverty/historical-income-people.html.

US Bureau of the Census. Historical Income Tables: People. Table P-8. Age–Black People, by Median Income and Sex: 1967 to 2020 (2021). https://www.census.gov/data/tables/time-series/demo/income-poverty/historical-income-people.html.

US Federal Works Agency. 1946. “Final Report on the WPA Program, 1935–1943.” General Services Administration. Accessed via The Library of Congress: https://www.loc.gov/item/47032199/.

West, Rachel, Rebecca Vallas, and Melissa Boteach. 2015. “A Subsidized Jobs Program for the 21st Century: Unlocking Labor-Market Opportunities for All Who Seek Work.” Center for American Progress. https://americanprogress.org/wp-content/uploads/2015/01/SubsidizedJobs-report3.pdf.

Wray, L. Randall, Flavia Dantes, Scott Fullwiler, Pavlina R. Tcherneva, Stephanie A. Kelton. 2018. “Public Service Employment: A Path to Full Employment.” Levy Economics Institute of Bard College. https://www.levyinstitute.org/pubs/rpr_4_18.pdf


  1. See Figure C in Austin, 2013. As the share of Hispanics within the white labor market data has increased, this demographic change has caused it to appear that there has been a small decline in the ratio. But when one limits the comparison to non-Hispanic whites, one sees little change. For example, averaging the unemployment rates for the first seven months of this year, the ratio is 1.8 including white Hispanics in the white data, but it is 2.0 excluding them. Authors’ analysis of Current Population Survey data from the US Bureau of Labor Statistics.
  2. For some of these proposals see Bashay and Banerjee, 2020; Austin, Glaeser, and Summers, 2018; Dutta-Gupta et al., 2018; Paul, Darity, and Hamilton, 2018; Wray et al., 2018; West, Vallas, and Boteach, 2015; Harvey, 2011; and Pavetti, Schott, and Lower-Basch, 2011.
  3. US Bureau of Labor Statistics, 2015. See Austin, 2022, for further explanation.
  4. See Austin, 2021b.
  5. This rough estimate is based on the following: The selected youth and prime-age jobs gaps by sex are multiplied by the Black Alone male and female median income for 2020 for the 15 to 24 year olds and the 35 to 44 year olds. The income data is from the US Census Bureau, 2021.
  6. Levy Yeyati, Montané, and Sartorio, 2019.
  7. The Georgetown Center on Poverty and Inequality has assembled a database with information on 51 subsidized employment programs. See Eckel et al, 2021.
  8. Pavetti, Schott, and Lower-Basch, 2011.
  9. West, Vallas, and Boteach, 2015.
  10. This calculation is based on Tables 8 and 25 in US Federal Works Agency, 1946.
  11. The 1940 Census recorded 12.8 million Black Americans. US Census Bureau, 1943. The 2021 Census estimate of the Black American population derived from the Census Bureau’s Quick Facts is 45.1 million, 3.5 times the 1940 population. https://www.census.gov/quickfacts/US.
  12. Cole, Jr., 1999, p. 7.
  13. For 2015 to 2019, the American Indian and Alaska Native population, including biracials, had a prime-age employment rate of 70.2 percent. The Black rate was 75.9 percent. Authors’ analysis of American Community Survey data from Ruggles et al., 2022.
  14. Authors’ analysis of American Community Survey data from Ruggles et al., 2022; Lowery, 2014.
  15. Bartik, 2022b. See also Center for American Progress, 2018.
  16. Bartik, 2022a.
  17. Authors’ analysis of American Community Survey data from Ruggles et al., 2022.
  18. Commuting zones are “multicounty areas that encompass most local commuting flows,” from Bartik, 2022a.
  19. To obtain this estimate, the prime-age jobs gap was multiplied by the 2021 median annual wage for Black Americans in Washington, DC. The wage data is from the State of Working XX dataset from the Economic Policy Institute.
  20. Authors’ analysis of American Community Survey data from Ruggles et al., 2022.
  21. Lowery, 2014.
  22. Authors’ calculations based on Census Bureau Table DP03 Selected Economic Characteristics, 2015 to 2019.
  23. Census Bureau Quick Facts: https://www.census.gov/quickfacts/fact/table/cartercountykentucky,US/PST045221.
  24. Authors’ calculations based on Census Bureau Table DP03 Selected Economic Characteristics, 2015 to 2019, and the State and County Employment and Wages from the 2021 Quarterly Census of Employment and Wages from the US Bureau of Labor Statistics.
  25. Authors’ analysis of American Community Survey data from Ruggles et al., 2022.
  26. Authors’ calculations based on American Community Survey data from Ruggles et al., 2022., and State and County Employment and Wages from the 2021 Quarterly Census of Employment and Wages from the US Bureau of Labor Statistics.
  27. Census Bureau Quick Facts: https://www.census.gov/quickfacts/fact/table/imperialcountycalifornia,US/IPE120220.
  28. Institute for Policy Research, 2020.
  29. Census Bureau Quick Facts: https://www.census.gov/quickfacts/fact/table/apachecountyarizona,US/PST045221.
  30. Authors’ calculations based on Census Bureau Table DP03 Selected Economic Characteristics, 2015 to 2019.
  31. Authors’ calculations based on Census Bureau Table DP03 Selected Economic Characteristics, 2015 to 2019, and the State and County Employment and Wages from the 2021 Quarterly Census of Employment and Wages from the US Bureau of Labor Statistics.
  32. Census Bureau Quick Facts: https://www.census.gov/quickfacts/fact/table/apachecountyarizona,US/PST045221.
  33. Harker, 2021.
  34. McCausland, 2022.
  35. US Federal Works Agency, 1946.
  36. Bartik, 2022a.
  37. This follows the recommendation in Bartik (2022a) for his Local Job Creation Block Grant.
  38. Dutta-Gupta et al., 2016.
  39. American Society of Civil Engineers, 2021.
  40. Endnotes omitted from quotation, Austin, 2021a.
  41. See for example, Harvey, 2011.
  42. For example, see the discussion of Minnesota’s MEED program in Bartik, 2022a, and the TANF Emergency Fund in Pavetti, Schott, and Lower-Basch, 2011.
  43. Bartik, 2022a.
  44. Gross, LeRoy, and Janis-Aparico, 2005.
  45. The Hill, 2019.
  46. Wray et al., 2018 argue that their design will not lead to a significant increase in inflation.
  47. Altman, 2022.
  48. Authors’ calculations based on American Community Survey data from Ruggles et al., 2022.
  49. The rough “back-of-the-envelope” job-creation estimate for the subsidized employment program is 3 million jobs. The estimate from Wray et al., 2018 is 15 million jobs.
  50. Wray et al., 2018 call for a $15 per hour wage and good benefits, but currently a third of workers have wages less than $15 per hour. Henderson, Kaitlyn. 2022. The benefits under the job guarantee proposal would likely attract even more workers.
  51. Horowitz (2020) discusses a program being tested in Austria, but it is restricted to “a long-suffering former industrial town.” In other words, it appears to be a targeted employment program for an area with high joblessness. There are calls for a job guarantee in Europe, however. For example, see Argitis and Koratzanis, 2021. The European Youth Guarantee provides “i) education and training for employment programmes; ii) remedial education school dropout measures; iii) labour market intermediation services; and iv) active labour market policies (ALMPs) aimed to affect labour demand, such as direct employment creation, hiring subsidies, and start-up incentives.” Thus, while the goal is to provide an intervention for jobless youth, that intervention can be things other than employment. See Escudero and López Mourelo, 2017.
  52. Hellerstein, Neumark, and McInerney, 2008.
  53. The Joint Center for Political and Economic Studies finds positive results for Black Americans in four of the six evaluations of training programs with data by race that they reviewed. But some of these positive evaluations were limited to educational outcomes; see Camardelle et al., 2022. If one only examines the two studies for which it is possible to compare employment outcomes, one, the MDRC evaluation, has a positive result and the other, the Public/Private Ventures evaluation, does not. Further, the MDRC evaluation shows a stronger positive employment result for Latinos. See Hendra et al., 2016; Maguire et al., 2010.
  54. Maguire et al., 2010.
  55. Hellerstein, Neumark, and McInerney, 2008.
  56. Quillian et al., 2017; Porter, 2021.
  57. Bivens, 2018; Levy Yeyati, Montané, and Sartorio, 2019.
  58. Austin, 2021b.
  59. Bivens, 2018; Baker and Bernstein, 2013.
  60. Quillian, Lincoln, Devah Pager, Ole Hexela, and Arnfinn H. Midtbøen, 2017; PNAS, 2017; Porter, 2021.

    Support Cepr

    If you value CEPR's work, support us by making a financial contribution.