A Path to Ending Poverty by Way of Ending Unemployment: A Federal Job Guarantee
Poverty in the United States, one of the world’s most wealthy and prosperous nations, is persistently high. Despite a complex array of social insurance programs in place, 43.1 million people remain in poverty. Because unemployment is a strong predictor of poverty, we propose a permanent federal job guarantee for all Americans. The program would provide full-time employment for any American over eighteen, offering at least nonpoverty wages plus benefits. Such a program will constitute a direct route to producing full employment by eradicating involuntary unemployment. It also will substantially increase worker bargaining power by removing the employer threat of unemployment. To make the case that the federal job guarantee is viable, this paper includes responses to five common criticisms lodged against programs of this type.
job guarantee, employment, poverty, human rights, second Bill of Rights
Poverty in the United States is persistently high even though the nation is one of the world’s wealthiest and most prosperous. Because an estimated 43.1 million Americans still live in poverty, some 13.5 percent of the population, conditions demand that the country take fresh action (Proctor, Semega, and Kollar 2016).1 Tools to alleviate the unnecessary suffering exist, [End Page 44] but they need to be activated (DeNavas-Walt and Proctor 2014). The United States does have a complicated array of social insurance programs in place that reach some of those in need: among them unemployment insurance, the Supplemental Nutrition Assistance Program (SNAP), Temporary Assistance for Needy Families (TANF), and disability benefits—both Supplementary Security Income and Social Security Disability Income.
Fighting poverty as a national priority is not a recent development. Fifty-two years ago, President Lyndon B. Johnson declared his War on Poverty, proclaiming that “many Americans live on the outskirts of hope—some because of their poverty, and some because of their color, and all too many because of both. Our task is to help replace their despair with opportunity” (1964). On the fiftieth anniversary of his declaration, the Council of Economic Advisors issued a progress report. The conclusion was that far too many Americans still experience poverty, in part because of “unemployment . . . inequality, wage stagnation, and a declining minimum wage” (2014).
The current social insurance regime has cut poverty rates nearly in half—reducing them from an estimated 27.3 percent for persons without government assistance to 15.3 percent after the programs mentioned are taken into account (Greenstein 2015). These programs certainly have reduced poverty, but the social insurance regime has shifted to a “work-based safety net,” providing the majority of assistance to the working poor—a group that would not exist under our proposed program in the first place. The changes in social insurance programs implemented under the Personal Responsibility and Work Opportunity Reconciliation Act (the 1996 welfare reform) resulted in a significant decline in government programs that provide benefits to people without earnings, contributing substantially to the increase in “deep poverty.”2
During the Great Recession, one of the largest holes in the current safety net became highly visible—the lack of support for the jobless. At the height of the crisis, conservative measures of the level of unemployment indicated that more than fifteen million Americans (10 percent of the labor force) were out of work (U.S. Bureau of Labor Statistics 2017).3 They were unable to make a contribution to traditional measures of economic productivity and economic growth, and they were struggling desperately to provide for themselves and their families. Broader measures of unemployment, such as U6, that better capture hardship during economic downturns indicate that 17.1 percent of workers were unemployed or working part time, despite wanting full-time employment.4
Figure 1 demonstrates just how large the gap became between the number of those seeking jobs and the number of jobs offered during the Great Recession. No amount of individual effort—hard work nor “pull yourself up by your own bootstraps” drive—could overcome the dramatic shortage of jobs available during the cyclical decline.
Moreover, unemployment is one of the strongest predictors of poverty, households whose usual breadwinners are out of work being three times more likely to be poor than working households (Achiron 2009, 13). But working households are not immune from the plague of poverty; a job in and of itself is not a sufficient condition to escape poverty. Given that at least 25 percent of workers earn wages below the poverty line (Mishel et al. 2012), and 44 percent of homeless individuals report having taken on paid employment in the past month (Burt et al. 1999), nonpoverty wages [End Page 45] need to be an essential component of reducing poverty.
Furthermore, the costs associated with unemployment go far beyond poverty. The nature of the harms from unemployment or underemployment are well documented. In addition to inflicting lasting damage on an individual’s labor market prospects, unemployment is associated with increased rates of physical and mental illness, alcohol and drug abuse, child and spouse abuse, failed relationships, suicide and attempted suicide, and a host of other personal and social ills (Goldsmith, Veum, and Darity 1997; Darity 2003).
Unemployment does not affect all groups equally; it varies greatly by race, as demonstrated in figure 2. Historical data indicate that unemployment rates for black workers are consistently twice those of white workers. This gap persists among groups with more education as well, with recent black college graduate unemployment at 9.4 percent, versus 3.7 percent among their white counterparts in 2016 (Bivens 2016).
In fact, the differential is so pronounced that there are many months when the Bureau of Labor Statistics reports that blacks with some college education have a higher unemployment rate than whites who never finished high school. Even when black students complete degrees in a statistics, technology, engineering, and mathematics (STEM) field, ostensibly fields in high demand by the labor market, they still experience markedly higher rates of unemployment. They also are more likely to end up in jobs that do not require a STEM degree (Jones and Schmitt 2014). Since 1972, unemployment has averaged double digits for black workers but has never fallen below 7 percent—a level reached only during times of economic crisis—for white workers.
But the ills of unemployment and poverty can be resolved by direct government action. In his 1944 State of the Union address, Franklin Delano Roosevelt introduced what he called an Economic Bill of Rights. The first “article” was a right to employment. In the absence of the provision of adequate opportunities for work by the private sector, demonstrated by the jobs gap [End Page 46] in figure 1, Roosevelt envisioned the maintenance of a public-sector option for employment for all. However, Roosevelt’s bold aim has not been realized, even if its ambition is embodied in the Full Employment and Balanced Growth Act of 1978.
This is unfortunate because a well-designed federal job guarantee (FJG) program would be a direct route to full employment and simultaneously eliminate involuntary unemployment and poverty in America.5 Such a program could be informed by and modeled after Great Depression–era projects such as the Works Progress Administration (WPA) and the Civilian Conservation Corps (CCC), as well as Argentina’s Jefes y Jefas and India’s National Rural Employment Guarantee (for Argentina, see Tcherneva and Wray 2005; for India, see Muralidharan, Niehaus, and Sukjtankar 2017). Any American wanting a job, at any time, would be able to obtain one through the public employment program.
The FJG would reach persons in the workforce who are subject to persistent exclusion from work, ensuring their capacity to secure employment. Groups continuously subjected to higher odds of joblessness, including exoffenders, recent military veterans, and racial-ethnic groups who experience discrimination, would be assured decent work at nonpoverty wages (Schmitt and Warner 2010; Loughran 2014; Darity 2003).6 These are the same groups subject to stubbornly high rates of poverty, in part because of their weaker job prospects (Proctor, Semega, and Kollar 2016; Western and Pettit 2010). The persistent racial unemployment gap, as discussed, would effectively be eliminated by the FJG.
By establishing a condition whereby the lowest paid job in the FJG program offers nonpoverty wages and benefits, including health insurance for the worker and their family, the federal job guarantee would set a new economy-wide floor on the level of compensation that the private sector would need to offer to attract workers. Minimum wage laws and living wage standards [End Page 47] are effective only to the extent that employees actually have jobs. A FJG, on the other hand, would combine an assurance of employment with an assurance of decent compensation.
The FJG also will function as an automatic stabilizer, its numbers of participants expanding during economic downturns and contracting during more prosperous times. Other articles in this double issue make important contributions on how best to improve and expand the existing system of social insurance. Our proposal offers a bold, yet historically grounded, alternative. The FJG fundamentally would alter the nature of poverty and the structure of the labor market by mandating a full employment economy achieved by the government taking the function of direct job creation.
Some of the authors have been advocates of a FJG for several years (Darity 2010, 2012; Darity and Hamilton 2012), but a number of issues and criticisms have been advanced about the FJG policy. Our objective is to provide a comprehensive design for a FJG that explicitly addresses these central criticisms and to further explore the impact of a FJG on poverty.7
We propose passage of legislation guaranteeing every American over the age of eighteen a job provided by the government via the formation of a National Investment Employment Corps (NIEC) (Darity 2010; Aja et al. 2013). The permanent establishment of the NIEC would eliminate persistent unemployment and poverty, ensuring that the United States is able to achieve full employment, as outlined by the Full Employment and Balanced Growth Act of 1978. Related legislation has already been introduced by U.S. Representative John Conyers (D-MI) in House Resolution 1000 (H.R. 1000).
This proposal moves beyond the incremental changes typically advanced as policy solutions for major social problems. Our proposal, decidedly non-incremental, aims to bring about pronounced structural change in the American economy. The FJG could simultaneously achieve the goals outlined in other proposals in this double issue (see Romich and Hill 2018; Dutta-Gupta et al. 2018). Rather than potentially operating de facto to subsidize low-quality jobs, our proposal effectively eliminates poverty for those willing and able to work by providing a guaranteed job at nonpoverty wages. This will set a floor for a decent standard of compensation in the labor market and fulfill the Humphrey-Hawkins full employment mandate.8 The proposal is predicated on the view that there is an absolute shortage of decent jobs, rather than there being insufficiently skilled workers to fill vacant positions in the private sector, as we demonstrated in figure 1.
Unlike other proposals in this double issue, many of which require people to first be exposed to poverty prior to gaining access to the benefits of the social safety net, our proposal is intended to preempt exposure to poverty altogether. Under the FJG, all workers seeking employment would be employed at the local, state, or federal level by the NIEC. The program would provide meaningful and remunerative employment across an array of public works projects addressing both the nation’s human and physical infrastructure needs.9
To simultaneously provide full employment and rid the United States of working poverty, workers would be paid at least $11.56 an hour. This wage rate would yield a salary of $24,036 a year at forty hours per week of year-round employment, equal to the poverty line for a family of four (DeNavas-Walt and Proctor 2014).10 This wage rate represents a minimum [End Page 48] entry level wage, but the program would be designed to enable workers to achieve promotion—and higher wages. We estimate a mean wage income for all employees of approximately $32,500.
In addition to wages, workers would be guaranteed a benefits package.11 We estimate that an additional $10,000 will be spent per worker per year to provide adequate health insurance and standard retirement benefits. This insurance would be comparable to current health insurance programs offered to federal employees and members of Congress. Beyond health insurance and retirement, employees would be guaranteed other benefits, including paid family and sick leave and one week paid vacation per three months worked. Again, we emphasize that compensation and benefits package will set a new floor in the labor market—compelling employers to provide competitive compensation packages, or risk losing workers to the FJG program.
Projects and employment under the program will be coordinated across the various levels of government—local, state, and federal—as well as the Indian Nations.12 At the federal level, we anticipate a wide array of major public investment activities, which may include fostering a transition to a green energy economy, extending access to high-speed rail, improvements in our public park service, revival and product diversification for the postal service, and an increase in general services across the economy.13
At the state level, we anticipate the states to undertake major infrastructure investment projects, as well as projects to improve the services they offer to their citizens. At the local level, we expect communities to undertake community development projects, provide universal daycare, maintain and upgrade their public school facilities, and improve and expand the services provided by their libraries.
In table 1, we provide three cost estimates for the proposal. Each estimate assumes the economy reaches full employment, by which we mean the elimination of both cyclical and structural unemployment. We assume U6 would be brought down to 1.5 percent.14 In column [End Page 49] one, we provide a snapshot of the FJG program’s costs drawing on experiences from the recent Great Recession. According to the Bureau of Labor and Statistics, the official peak unemployment rate during the Great Recession was 10 percent (15.3 million Americans).
We believe these figures grossly underestimate the numbers of workers in need during the Great Recession. Considering a broader measure of unemployment (U6) provides us with a more reasonable estimate for the number of workers that may seek employment through the FJG. During the peak of the crisis, U6 reached 17.1 percent. Removing those under the age of eighteen, since minors would not qualify for the program, we find that an estimated 24.1 million workers might have sought employment under the program. However, this might be an unrealistically high estimate of expenses, if the FJG has been in place prior to the recession. Due to its buffer stock mechanism, employment shocks would have been moderated and full employment would have been maintained.
Next, we also estimate two scenarios under recent economic conditions using July 2016 employment data. The first scenario assumes modest uptake under the FJG, on the assumption that all workers currently counted in U6 engage in employment through the National Investment Employment Corps (NEIC). In this case, 13.2 million workers may seek employment, demanding 11.9 million full-time equivalent jobs. The gross cost of the program, including [End Page 50] benefits and materials, would be $654.6 billion.
The second case based on recent labor market conditions estimates program uptake and expenditures assuming workers earning wages below the minimum offered by the NEIC also would partake in the FJG. Although we believe U6 provides a reasonable approximation of program uptake, we recognize that the establishment of a FJG will transform much of the labor market. In turn, we analyze a third estimate that represents an upper bound for uptake and cost under a FJG. This estimate assumes that all workers currently earning below the base wage offered by the NEIC (about one-quarter of all employed workers according to Bureau of Labor Statistics data), plus workers counted under U5, will seek employment with the NEIC.15 Under this scenario, program costs would increase sizably.16
Although these are large initial employment and cost projections, the macroeconomic stimulus effects from such a program would be substantial, generating significant employment in the private sector, thereby mitigating workers’ demand for public-sector employment—and contributing to cost containment. Philip Harvey estimated the indirect job creation effect of a government direct jobs program, similar to the FJG, and calculated that, for every directly created job by the government, 0.26 indirect jobs would be created through the private sector (2011).17
Costs associated with the FJG only represent half of the equation. The initial cost of the FJG would be offset by significant cost savings through reducing enrollment in many existing federal and state social insurance programs, by maintaining state and municipal tax bases, by increasing the growth rate of gross domestic product (GDP), and by substantial productivity and capacity gains in the U.S. economy.
The Congressional Budget Office’s baseline projections provide cost estimates for the major social insurance programs over the next decade, from 2016–2026, which we present in table 2 (Congressional Budget Office 2016). Following the successful rollout of the FJG, the number of beneficiaries eligible for SNAP is likely to fall substantially. Unemployment insurance, which peaked at $162.5 billion in 2010, would be reduced significantly because workers would have an option to obtain employment through the FJG (see Congressional Budget Office 2012).18 TANF also could be nearly eliminated [End Page 51] because the FJG would fill the gap for those in need.19
In addition, fewer families would likely qualify for the existing Earned Income Tax Credit (EITC) plan.20 For instance, those working without dependents would no longer be eligible, as their income exceeds the threshold to qualify for the EITC. Those with dependents would likely see a decrease in their EITC because the level of compensation under the FJG will place most households beyond peak benefits. Through the reduction of these programs, coupled with the economic returns from investment under the program such as green energy and infrastructure, we believe a substantial portion of the costs of the FJG will be offset; hence, the net additional expenditures required for the program would be considerably less than the total costs estimates reported in table 1.21
On a per-dollar basis, the FJG would be more effective in creating jobs than the indirect incentive effects of stimulus measures, like those pursued under the American Recovery and Re-investment Act (ARRA) of 2009. The ARRA came to the tune of $787 billion. The Economic Stimulus Act of 2008 added another $170 billion. Alan S. Blinder and Mark Zandi add the total bill for fiscal stimulus in the government’s attempt to curb the recession, estimating a grand total of $1.067 trillion (2010).22 The indirect job creation costs through the ARRA is estimated at $100,000 spent for each full-time job year (Dube, Kaplan, and Zipperer 2014), which is [End Page 52] substantially higher than the costs we estimate for direct job creation, which also would generate indirect job creation stimulus (for an in-depth discussion of the job efficiency of direct government job finance as opposed to other forms of government stimulus, see Harvey 2011).
Under the FJG, the money would primarily flow into the hands of those in need. This was not necessarily the case under the responses to the Great Recession. For example, Juan Montecino and Gerald Epstein find that quantitative easing actually exacerbated income inequality through equity price appreciation (2015). In contrast, the FJG affords a direct countercyclical approach to smooth business cycles, eliminate poverty-wage employment, and boost long-term growth.
Although benefits effectively will disproportionately aid the poor and those struggling to find private-sector employment, the private sector could see a boost as well. Because the poor have a high marginal propensity to consume, aggregate demand would be maintained—or even rise, leading to increased demand for private-sector goods. Plus, the countercyclical effects of the FJG will help maintain aggregate demand during future recessions. Also, the improved physical and human infrastructure efficiency resulting from the program will facilitate productivity gains for the private sector as well.
Workers under the NIEC will be able to acquire the necessary skills, opportunities for advancement, on-the-job training, and professional experience to aid in long-term career development. These advantages to workers will be provided in part through a training provision under the FJG. As shown by the CCC, the WPA, Argentina’s Plan Jefes, and India’s National Rural Employment Guarantee, even low-skilled workers can be assigned to valuable work in a relatively short time. Although some jobs in the FJG will rely on basic labor, others will require workers with additional skills. Through a federally maintained jobs bank, projects will be matched with the skills of the local workers in need of employment. This is similar to the content of Representative Conyers’ bill (H.R. 1000) and will minimize any potential skills mismatch issues.
Because workers have the option to freely enter or exit the FJG, employers who pay below the nonpoverty wage established by the FJG and fail to offer competitive benefits will not be able to attract employees, except perhaps those interested in part-time or temporary work, or motivated to work in such jobs for less pecuniary reasons. Therefore, the FJG acts as an effective wage floor, reducing economic inequality, especially at the low end of the income distribution.
The transformative nature of this proposal on the labor market should not be underestimated. Productivity and the real minimum wage once rose in tandem, but have diverged since the 1970s (Cooper 2015). By functioning as an employer of last resort, the government greatly improves the bargaining power and fall-back position of workers in general by removing the threat of unemployment and effectively eliminating involuntary unemployment.
By providing a job guarantee, the proposal has the added advantage of greatly reducing the unemployment, underemployment, and poverty of permanently stigmatized groups, which are subject to discriminatory exclusion from employment opportunities. For instance, field experiments conducted by Devah Pager in Milwaukee, Wisconsin, and in New York City reveal that among males of comparable ages and levels of education, white males with criminal records were more likely to get callbacks for jobs than black males with no criminal record (Pager 2008). It is noteworthy that the use of criminal background checks is outlawed for most jobs in Milwaukee; nonetheless, Pager finds that workers signaling prior incarceration were still substantially less likely to receive call-backs there.
debunking common criticisms
The FJG is one of many potential poverty alleviation programs; however, we believe the FJG offers unique payoffs by working toward building a just and inclusive economy through a full employment economy. To ensure that the FJG is a viable program, in this section we respond to five of the common criticisms lodged against programs of this type. What type of work will people actually do? Why not just adopt a universal basic income guarantee (UBI)? How do [End Page 53] we elicit effort if people cannot be fired? Would such a program be prohibitively expensive? Would the program hurt small businesses? Each of these is addressed in turn.
What Type of Work Will People Actually Do?
What type of socially useful work could the members of the NIEC engage in? During the Great Depression, the WPA, created by Executive Order 7034 by President Franklin Roosevelt, employed more than 8.5 million workers from 1935 to 1943. In November 1938, at peak program size, it employed 3.3 million Americans, an estimated 6.5 percent of the nation’s labor force (Hansan 2013). The program assisted these hardworking Americans during times of distress, providing them with the dignity of work and the ability to put food on the table and meet necessary bills. How much did these workers accomplish? Just to name a few: 650,000 miles of new or improved roads; 124,000 new or improved bridges; 39,000 schools built, improved, or repaired; 85,000 public buildings built, improved, or repaired (excluding schools); 8,000 new or improved parks; 4,000 new or improved utility plants; 16,000 miles of water lines installed; 950 airports or airfields built, improved, or repaired; 1,500 nursery schools operated; 2,300 personal accounts of slavery gathered; 225,000 concerts performed.
Similar to their counterparts in the WPA, workers in the NIEC could repair, maintain, and build the nation’s deteriorating infrastructure, retrofit our buildings—aiding in a green energy transition, saving homeowners and renters money, reducing our carbon footprint—provide free or low-cost, high-quality preschool and after-school services, function as teacher’s assistants in the classroom, engage in community development projects, reinvest in our nation’s parks, rejuvenate our defunded postal service, as well as perform other socially and economically rewarding tasks.
In 2013, the American Society of Civil Engineers released their Report Card for America’s Infrastructure, giving the country’s crumbling infrastructure a D+ grade (ASCE 2013). More recently, the ASCE reported that the U.S. economy needs $3.32 trillion in funding to address its infrastructure gap and ameliorate public safety concerns, $1.44 trillion of which is not currently funded. They project that failure to act on America’s crumbling infrastructure would result in $7 trillion in lost business sales by 2025 (ASCE 2016). Furthermore, they estimate that if the infrastructure gap is not closed, between 2016 and 2025 it could cost the U.S. economy $3.9 trillion in lost GDP, $7 trillion in lost business sales, and 2.5 million jobs. Overall, they conclude that the average household will lose $3,400 in disposable income per year if the deficiency remains unaddressed.
Economists have expressed concern about the slowdown in the growth of productivity in our economy (Syverson 2016), yet we know that large-scale investment in infrastructure—both physical and human—can have a measurable impact on capacity by increasing available resources and enhancing the productivity of existing resources (Munnell 1992; Heintz, Pollin, and Garrett-Peltier 2009). The FJG would be able to adequately address these needs and more, alleviating these costly and unnecessary shortcomings in the economy.
To make the best use of the labor available through the FJG program, states, counties, and municipalities can conduct an inventory of their needs and develop a jobs bank. The program could give priority to the most urgent projects to aid the most distressed communities. Although we highlight some investment opportunities that are needed now—such as retrofitting our buildings and heavily investing in our infrastructure—the jobs bank, managed through the NIEC, would function as a constantly updating dynamic entity, shifting with local, state, and federal needs.23
Under the FJG, many additional services would be provided to Americans, resulting in an increase in discretionary income and improvements in quality of life. Some of those services will address the nation’s human infrastructure needs. These will include the provision [End Page 54] for childcare and eldercare.24 Some care providers will be trained to deliver services to persons with special needs. The current exorbitant cost of childcare is a major obstacle for many parents, restricting their access to the labor market and greatly exacerbating inequality in access and opportunity for children. A recent report from the Economic Policy Institute, found that childcare is more expensive than a college education in many states (Bivens 2016). Robert Lynch and Kavya Vaghul estimate that universal high-quality pre–K education would yield an estimated $10 billion annually in benefits from 2016 to 2050 while providing greater access to high-quality educational services (2015).25 Solidifying high-quality universal pre–K education, childcare, adult care, and elder care across the United States would all result from a federal job guarantee that includes provisions for human infrastructure investment. Although our proposal is facially gender neutral, it has basic elements that could disproportionately benefit women. This would greatly ease time and financial burdens borne disproportionately by women.
Some opponents of the FJG argue that if the market is not currently providing these jobs, then the government has no reason to do so. For example, according to Guy Standing, “a job guarantee is a form of subsidy, in that it involves a payment for doing something for which there is no proven market demand. Unlike a universal unconditional transfer, all labour subsidies involve both substitution effects and deadweight effects” (Standing 2013).
In our economy, we can find numerous examples of unfulfilled voids by the private sector where the government has had to step in to provide necessary and socially beneficial services through additional employment and investment. The core purpose of federal taxation and expenditure is to provide the American people with the government services and public goods.
For instance, prior to the American Civil War, the United States did not have a government-run public fire department (Tebeau 2012). Instead, fire departments were private or organized solely on a volunteer basis, at times leaving poor houses and neighborhoods to burn, and the well-to-do were privy to what is today a public good with potential spillover calamity—adequate fire protection. Without government spending, production of public goods will be inadequate, resulting in socially inefficient outcomes.
From our perspective, the government has a public goods and equity role to fill gaps where the market fails—and there are plenty of cracks and canyons to be addressed. We believe that a wide range of socially useful jobs can be filled by the ranks of the unemployed with the assistance and coordination of the NIEC. After all, children are undereducated, too few have adequate medical care, greater care and service is needed for our elderly, our parks are under-staffed and underutilized, and our nation’s transportation infrastructure is inadequate. Some of these jobs can be countercyclical, like infrastructure investment, and others can function as a permanent expansion of government services, such as universal preschool.
Why Not Just Adopt a Basic Income Guarantee?
The universal basic income proposal has gained followers across the political spectrum as a viable path to fight increasing deprivation; it also is a non-incremental, bold policy that merits comparison with the FJG. The FJG and the UBI are not necessarily mutually exclusive. In the end, it depends on the structure of the specific policies proposed: is UBI posited as a substitute or complement to the FJG? [End Page 55]
Although we oppose the idea of substituting a UBI for a FJG, we believe some form of a UBI, such as a negative income tax, could be beneficial if coupled with the FJG, to ensure an adequate standard of living for all.
We argue in favor of FJG over the UBI on the following grounds:
the FJG provides the dignity of nonpoverty employment for all who seek it;
the FJG enables the nation to fulfill a host of socially useful tasks that are not currently provided, or are underprovided, by the public sector;
the FJG carries a lower inflation risk than the UBI;
the FJG contributes directly to macroeconomic stabilization; and,
perhaps surprisingly, the FJG will cost considerably less.
Advocates of the UBI have been critical of the FJG proposal (Van Parijs 1995; Standing 2002; Standing 2013). To justify a UBI, Standing fundamentally questions the ability of a market society to provide jobs for all, claiming that market societies rely on restricted job openings to discipline the poor; however, with the presence of a FJG, full employment is achieved. Standing’s objection rests on a philosophical opposition to the nature of work in modern society. From his perspective, the requirement to work is fundamentally punitive and unjust. Additionally, Guy Standing claims that the FJG would condemn the poor to have to work. On the contrary, the work that we envision via a FJG provides the dignity of contributing to social welfare in a social setting, which in turn is associated with economic, physical, and mental well-being (Goldsmith, Veum, and Darity 1997; Darity 2003).
Workers employed under the FJG would enable the nation to fulfill a host of socially useful tasks either not provided or inadequately provided by the private sector, including those outlined. Today, we still observe the enduring benefits of projects constructed under the WPA and CCC—the FJG will ensure that similar projects are constructed, providing economic, social, and environmental benefits for generations to come.
Unlike the UBI, the FJG generates productive activity via the public provision of goods and services, lowering the inflationary pressures associated with the allocation of similar levels of income through a UBI. The FJG may have some inflationary pressure, but Pavlina Tcherneva has expressed concern that the magnitude of a UBI might even lead to hyperinflation. Inflation would dampen the real effects of a UBI program. This problem would be exacerbated if workers receiving the UBI exited the workforce altogether and reduced output via voluntary unemployment (Tcherneva 2013).
A substantial benefit of the FJG is that it functions as a strong automatic stabilizer in the economy, expanding during economic downturns, and contracting during economic booms when the private sector’s demand for labor increases and workers migrate from the FJG program into the private sector. This mechanism provides substantial macroeconomic stabilizing effects on the economy—potentially reducing the magnitude and frequency of economic downturns. Regarding the UBI, no countercyclical measures are built into the model—allowing financial markets and businesses cycles to continue causing unnecessary economic hardship and job loss for Americans (Goldsmith, Veum, and Darity 1997; Darity 2003).
To compare the costs of the two programs, we need to identify the level and ways a UBI could be funded, because levels of generosity could vary greatly. Reviewing work from various UBI advocates, we see that the following three features hold throughout: the policy is universal, the UBI is distributed to individuals, and the UBI is set at a level to support a basic living standard, with a goal of poverty elimination (see Van Parijs, 1995; Standing 2002; Standing 2013).
Estimates for the UBI vary, but Charles Clark estimates the UBI would have cost roughly $1.98 trillion in 1999, equivalent to $2.86 trillion in 2016 dollars (2003, 150). Harvey also calculated the total cost of a UBI, finding at the time of his study that it would cost $2.23 trillion, equivalent to $2.98 trillion in 2016 dollars (2005). Both of these calculations are based on a payout of approximately $12,500 per person per year. These are annual cost estimates—estimates that represent a cost more than double that of the FJG during the most severe recession [End Page 56] in almost a century (see table 1), and more than four times our estimates under current economic conditions with a modest uptake assumption.26
These estimates showing that the FJG costs are much less than the UBI do not account for the returns to investments that would take place under the FJG—resulting in substantial GDP, employment, and productivity growth. Finally, the costs associated with UBI will not ebb and flow with the business cycle, hence it lacks economic stabilization properties. For these reasons, we strongly support a FJG when UBI it is posited as a substitute, though we recognize that the two can function as complements.27
How Do We Elicit Effort if People Cannot Be Fired?
The federal job guarantee ensures that workers cannot be fired and left without alternative employment. According to Janet Yellen, writing two decades before her appointment as chair of the Federal Reserve, if full employment were to be achieved, a worker would automatically shirk since they would be ensured employment at another firm (1985). But economic history indicates that such a claim is highly exaggerated, as strong labor markets have coexisted with high levels of economic growth (Bivens 2016).
If there were widespread worker shirking under conditions of strong labor markets, such strong periods of growth are unlikely to have occurred. Other studies, like those analyzing care work (England, Budig, and Folbre 2002) and management practices (Brockner et al. 1992), acknowledge that pecuniary incentives alone cannot adequately explain variations in worker effort.
Eliciting worker effort can be achieved through a variety of mechanisms. For instance, Samuel Bowles argues that from “a social efficiency standpoint . . . [m]ore carrot and less stick would affect a technical efficiency improvement” (2012, 279). If history is any guide, we can look to the accomplishments under the CCC and WPA, programs that were highly effective in providing socially and economically beneficial employment.
President Roosevelt and WPA administrator Harry Hopkins took great measures to minimize shirking or corruption throughout the WPA by establishing the “Division of Progress Investigation” (Wallis, Fisherback, and Kantor 2007). According to Paul Krugman, “this program was so effective, that a later congressional investigation couldn’t find a single serious irregularity it had overlooked” (2007, 62). This strategy may have been effective during a severe national crisis, but additional mechanisms can be put in place to minimize potential shirking under the FJG.
First, the FJG program will have an oversight committee, similar to the Division of Progress Investigation under the WPA. These positions can be filled by FJG employees, adding no additional cost to the program. In addition, wage variation will be built into the program. The prospect of promotions within the public-sector employment job ladder scheme provide financial incentives for workers to perform on the job. Although minimum compensation is set at the poverty line, we anticipate a mean salary that is 35 percent above the minimum, allowing for promotions, including higher compensation, for workers who develop or possess more specialized skills.
Finally, compensation will depend on the worker’s actually showing up for the duration of the work day. Workers under the FJG are paid hourly; thus, if a worker fails to show, wages will not be paid.28
In the case of the FJG, much of the work undertaken will be local, community-based projects. Because many of the projects are embedded in workers’ communities, workers themselves will be more vested in these [End Page 57] community-based projects, as occurred in the direct employment program in Argentina (Tcherneva and Wray 2005). Thus, through localization, reward systems, and proper oversight under the FJG program, we do not anticipate worker effort to be a substantial obstacle.
Would Such a Program Be Prohibitively Expensive?
Although the initial financial cost estimated in table 1 may at first glance appear high, the initial costs will be offset by extensive cost savings through reduction of current social insurance programs, many of which would be substantively less necessary, return on investment within the program itself, moderation of economic downturns, expansion of economic capacity, and increases in tax revenues, particularly at the local and state levels. The latter entities frequently are bound by fiscal year constraints to balance their budgets that lead to tremendous cutbacks in the services they provide during economic downturns.
The FJG substantially will reduce the need for the current levels of social insurance programs as individuals become ineligible for benefits such as SNAP after their earnings surpass the designated thresholds. In addition, progressive taxation, as outlined, could be enacted if politicians sought to offset the remainder of the FJG. Because the funding of the FJG will have Keynesian stimulus effects even as it directly provides full employment, we expect that increases in economic production will help reduce net costs further.
Part of the initial cost of the FJG is due to the failure of the U.S. government to provide universal health care for its citizens. If the government were to provide all Americans with health-care coverage, program costs would be reduced considerably.
Would the Federal Job Guarantee Hurt Small Businesses?
The goal of the FJG is to simultaneously eliminate poverty and unemployment while supporting a robust and inclusive economy—including a small-business sector. The FJG will have adverse effects on small businesses that rely upon low-wage labor. But the magnitude of these effects can be approximated by looking both at the historic data and recent research on the effects of increases in the minimum wage.
Under the FJG program, in 2016, the wage would have been $11.56 plus benefits. An increase of this magnitude would increase the wages of millions of workers—ensuring all the dignity of a decent wage.29 This degree of compensation is not beyond historical trends tracking the magnitude of the minimum wage. During the late 1960s, when we saw peak real values of the minimum wage, the small-business sector did not collapse—nor do we anticipate it doing so under the FJG. Rather, we do anticipate, initially, a redistribution of profits in favor of labor and a modest rise in price levels.30
Recent research on the minimum wage provides additional evidence. Starting with work by David Card and Alan Krueger (1994, 2000), we now have compelling evidence on the employment effects of higher minimum wages. Their approach has been refined and replicated by Arindrajit Dube, William Lester, and Michael Reich (2010). Consistently, these researchers find no evidence of job losses in high-impact sectors from modest increases in the minimum wage.31 However, larger employment effects may take place since the total compensation under the program—inclusive of benefits—amounts to $16 an hour, an amount beyond the scope of existing empirical minimum wage [End Page 58] models (for additional discussion of the minimum wage, see Romich and Hill 2018).
Some employment effects could be overcome by growing evidence indicating that a rise in the minimum wage can reduce job turnover (Dube, Kaplan, and Zipperer 2014) and increase per capita output to the extent that higher wages spur greater productivity (Reich et al. 2016). Reductions in turnover and increases in productivity explain why many small businesses have chosen to invest in higher compensation packages for employees with great success (Ton 2012) and generally support (60 percent) a $12 minimum wage pegged to inflation (Small Business Majority 2015).
Finally, through the FJG, we expect to see a substantial rise in effective demand, as well as investment in infrastructure across the country—boosting sales while driving down transportation and utility costs for small businesses. Furthermore, research has found that in many instances higher minimum wages are associated with superior outcomes for small businesses. A study by the Fiscal Policy Institute analyzed the impact of higher minimum wages on small businesses between 1998 and 2001, finding that small businesses grew twice as fast in states with higher minimum wages—3.1 percent to 1.6 percent (2004).32
Because workers, especially those at the low end of the income distribution, have a higher marginal propensity to consume, we would expect a substantial uptick in sales for businesses and perhaps an uptick in small businesses in currently economically depressed geographical areas. A report by the Chicago Federal Reserve Bank on the impact of spending as a result of a minimum wage increase found every dollar increase in the wage resulted in an annual spending increase of $2,996 (Aaronson, Agarwal, and French 2007).33 Given the robust literature on the minimum wage, we do not believe the FJG will drastically reduce the small businesses sector.
discussion and conclusion
Not only does the idea of a federal job guarantee have a lengthy history in American politics, it also has significant political support. For example, the Black Youth Project 100’s “Agenda to Build Black Futures” explicitly includes the charge that “All adults who want a job should have a right to employment through public or private opportunities through a federal jobs program.”
In 2014, Jesse Myerson’s article in Rolling Stone proposed five policies that should be supported by the millennial generation; one of them was a government employment guarantee. When the Huffington Post commissioned a national survey to assess the degree of support for each of the five, the only one with substantial support was the job guarantee (Swanson 2014; Resnikoff 2014). Overall, a plurality of respondents, 47 percent, said that they favored a job guarantee versus 41 percent who said no and 12 percent who had no opinion. Fifty-nine percent of households with incomes less than $40,000 favored the job guarantee, and 36 percent of households with incomes greater than $100,000 favored it.34
Given this public support, perhaps at the [End Page 59] least, a local experiment is warranted—a local job guarantee. To help workers and communities with the most need, the local job guarantee would cover an economically depressed county or city and be coupled with residency requirements to eliminate concerns of in-migration. Partnering with a local government to implement a jobs guarantee program could lift thousands out of poverty and revitalize a regional economy.
The program would work with government and local community and business groups to assess sectors most in need of public services, such as local infrastructure or social services. In general, the local guarantee would mirror the proposed federal job guarantee; thus, the jobs are meant to provide local services and public goods in the region, rather than to produce private goods that can be exported out of the region. However, by providing public services and goods, the job guarantee would no doubt boost private economic productivity as well.
Such a local guarantee would test the economic and poverty-alleviating benefits of a job guarantee program, and it can also serve as a pilot program for a larger scale jobs guarantee program by providing improved estimates of costs and participation rates in relation to pre-guarantee conditions. As a pilot program, it would demonstrate how a job guarantee would work in practice in the United States, providing evidence to extrapolate toward the proposed federal version.35 In addition, as demonstrated by the Fight for 15, a minimum wage campaign that has succeeded in New York and California, and health-care reform, when the national Affordable Care Act was modeled after Massachusetts’s reforms, having successful local pilot programs can go a long way toward making a federal version politically feasible.
Mark Paul is a postdoctoral associate at the Samuel DuBois Cook Center on Social Equity at Duke University.
William Darity Jr. is Samuel DuBois Cook Professor of Public Policy, African and African American Studies, and Economics, and director of the Samuel DuBois Cook Center on Social Equity at Duke University.
Darrick Hamilton is associate professor of economics and urban policy at the Milano School of International Affairs, Management and Urban Policy and Department of Economics, New School for Social Research, and director of the doctoral program in public and urban policy at the New School.
Khaing Zaw is a research associate at the Samuel DuBois Cook Center on Social Equity at Duke University.
1. The aggregation of the national poverty rate is too general to describe the scope of poverty for certain groups. For instance, the black poverty rate of 24.1 percent is more than twice the white poverty rate of 11.6 percent.
2. Deep poverty is defined as an individual or household with income that is below half of the poverty line (for further discussion on changes in extreme poverty following the 1996 welfare reform, see Shaefer and Edin 2013).
3. U3 is the International Labour Organization official unemployment rate that includes individuals that are unemployed and have actively looked for work within the past four weeks.
4. U6 is a broader unemployment, or underemployment rate, which, in addition to U3, includes “discouraged workers,” or those who have stopped looking for work due to current economic conditions; other marginally attached workers who are willing and able to work but have not actively sought employment in the past four weeks; and part-time workers who seek but cannot attain full-time employment.
5. Although many economists have abandoned the notion of full employment, we follow Keynes’s definition—the elimination of involuntary unemployment. Under the FJG, the full employment unemployment rate would near 1.5 percent, representing short-term frictional unemployment.
6. The nonpoverty wage examined in this paper is $11.56 an hour, equal to the poverty line for a family of four. The choice of this rate as the base wage is elaborated on later in this article.
8. The Humphrey-Hawkins Act, also known as the 1978 Full Employment and Balanced Growth Act (PL 523. Full Employment and Balanced Growth Act of 1978, 95th Cong., 1978), established the full employment portion of the Federal Reserve’s dual mandate.
9. The inclusion of human infrastructure refers to services such as child care, elder care, and other services that facilitate improving of people’s life quality and personal development.
10. Rather than setting a wage that is means-tested or linked to household size, the program sets an absolute minimum wage based on an above poverty threshold for a family of four. The program does not presume that all families will have one working adult. Such an approach does not directly link wages to household size or composition. In some cases, there may be some families with one adult working in the private sector and another under the auspices of the federal job guarantee. Our anchor on a family of four is consistent with the prevailing social norms regarding family size (see Gao 2015).
The job guarantee is a guarantee per adult, not per family, and hence a family of four with two adults working in the program would have a wage commensurate with twice the poverty line. We recognize that the minimum wage would not be adequate to lift every family above the poverty line. For instance, the wage would not lift a single-headed family with four children above poverty.
11. If the United States were to pass a public option for health insurance, such as a Medicare-for-All type program, the cost of the FJG would be reduced substantially.
12. Coordination between government entities would be necessary to achieve successful implementation of the program. We recommend a similar structure to that undertaken by the WPA. In additional to direct federal hiring, the secretary of labor could administer “employment grants” enabling eligible entities to also engage in direct employment projects. These projects should be designed to address community needs and provide socially beneficial goods and services to communities and society at large. There would be an oversight board at the national level to ensure local capacity (available workers, knowledge, staff, and funding for materials) could successfully implement the proposed project. In addition, we recommend that secretary of labor work with federal agencies to identify areas of needed investment in the U.S. economy that would not displace work planned by existing government agencies. If projects at the local, county, or state level are not enough to maintain full employment in the region the secretary of labor will intervene in the locality to provide adequate employment opportunities.
13. Robert Pollin and his co-authors at the Political Economy Research Institute estimate that a transition to a green energy economy would amount to an estimated $200 billion annually. This largely could be undertaken by workers in the FJG program (2014).
15. Similarly, this estimate brings the unemployment rate down to 1.5 percent. We use the U5 unemployment rate in this calculation because we do not want to double count workers who are currently in part-time employment.
16. Note that similar to the findings of the extensive literature on the minimum wage, we do not anticipate that setting a reasonable floor in the labor market will result in complete private-sector crowding out.
17. A classic concern with the FJG is that a substantial exodus from the private sector will occur (for example, private-sector employment crowding). A few points are notable. First, the private sector jobs that are most vulnerable to competition from a FJG are generally the least desirable to workers, offering them the lowest wages and benefits package. Second, the minimum level of compensation offered in our proposal is not out of line with historic minimum wages once inflation and productivity increases are taken into account. In fact, if the minimum wage increased with inflation and productivity from its peak in 1968, it would have been $18.70 per hour in 2016 (Cooper 2015). Minimum wage studies have not uncovered evidence of mass layoffs in low-wage sectors following substantial minimum wage hikes. In part, the change in the floor of minimum compensation induced by a FJG would redistribute some income from profits to wages. This will likely lead to higher prices in some sectors, and to some firms exiting from industries that offer the lowest wages and compensation packages (for a review of the magnitude of potential price increases under a $15 minimum wage in the United States for the fast-food industry, see Pollin and Wicks-Lim 2016).
18. Once the FJG is established and running at full capacity, the government may decide to reduce the duration of unemployment insurance. When the Committee on Economic Security provided its report to President Roosevelt on unemployment insurance and a job guarantee, the committee suggested unemployment insurance be capped at fourteen or fifteen weeks (Committee on Economic Security 1935). Such an adjustment seems reasonable.
19. The FJG effectively eliminates the notion of the working poor. Thus, if a single parent with one or two dependents were working full time under the FJG, which would be possible in part due to low-cost childcare under the program, the household’s income would be above the poverty line, and above the TANF eligibility criteria.
20. Some individuals previously unemployed or outside the labor force may receive EITC benefits once employed through the FJG. Anyone employed in the FJG full time would be in the phase-out stage of EITC benefits due to the nonpoverty wage. As pointed out to us by an anonymous referee, all childless workers and two-earner families with one child that are working full-time under the program would not receive any EITC benefits. However, previously unemployed families with one earner and children would receive EITC benefits. Also, families with two full-time earners and multiple children, who had previously been unemployed, would also receive EITC benefits. In addition, as the anonymous referee pointed out, many part-time minimum and low-wage workers who had been on the phase-in range of the EITC would move to higher EITC payments on the FJG.
On net, whether the FJG will induce higher or lower EITC costs depends on whether the cost reductions from previous recipients phasing out of EITC exceeds or falls below the additional EITC costs associated with the uptake from previously unemployed single workers with children, two-earner families with multiple children (newly eligible EITC families), and from part-time and low-wage workers whose FJG participation could yield higher EITC benefits. Nonetheless, as mentioned, the FJG would be consistent with federal fiscal cost savings from reductions in unemployment compensation expenses, and in other federal anti-poverty programs such as TANF and SNAP, due to FJG participation.
21. Congress also could enact tax reform to address cost concerns. A modest financial transaction tax could generate $340 billion per year alone (Pollin, Heintz, and Herndon 2016). Other taxes, such as a tax on carbon, modest estate and gift taxes, tax reform on capital income, and a reasonable minimum tax on foreign earnings could also be enacted.
22. Their estimate does not include the additional stimulus measures undertaken by the Federal Reserve.
23. Some have questioned the government’s ability to manage such a program, but if history is a guide such concern is overstated. For instance, the Civil Works Administration, which ran from 1933 to 1934, was fully functional within two months and employed more than 8 percent of the U.S. workforce (Harvey 1989).
24. Noteworthy is that under previous direct employment programs administered by the U.S. government, large-scale childhood education projects were administered, including the employment of more than thirty-three thousand teachers during the WPA to offer educational programs, including program-operated preschools (Harvey 1989).
25. Other scholars have put forth alternative proposals for determining the types of jobs to be undertaken. Pavlina Tcherneva argues for a “social enterprise model” for job creation within a FJG program (2012). She advocates a federal job guarantee initiative that engages heavily with the nonprofit sector, including existing and emerging nonprofit enterprises, to place workers in employment with these organizations.
26. Even in our extreme high-end estimate the cost of the job guarantee program would only amount to $2.14 trillion, substantially less than the cost of the UBI.
27. Other scholars, such as Dorian Warren, have discussed modified basic income proposals, which incorporate a progressive distribution scheme. His plan for a Universal PLUS Basic Income includes a form of reparations, providing black citizens with additional resources to adjust for the unjust treatment they have received (Warren 2017).
28. Exceptions include the use of benefits provided under the program, such as medical leave and family leave.
29. One avenue for reduction of the wage bill for small businesses, and program costs for the FJG, would be to institute a universal health care program in the United States. Instead of coupling a basic human right like health care to employment, decoupling these would substantially reduce financial burdens on small businesses, and increase worker productivity.
30. In a study of employment effects for the fast food industry, if a $15 minimum wage were to be implemented, Robert Pollin and Jeannette Wicks-Lim find that the fast food industry could absorb wage increases without cutting their employment levels by reductions in turnover and “modest annual price increases” (2016).
31. Sectors typically affected by increases in the minimum wage include retail trade, health care and social assistance, and restaurants. Although some studies, such as those by David Neumark and William Wascher (1992) and Neumark, Ian Salas, and Wascher (2014) find more sizable job losses from an increase in the minimum wage, we believe that a recent paper by Sylvia Allegretto and her colleagues effectively refutes those findings (2016). In the paper, the authors properly account for existing state level trends to assess employment effects of raising the minimum wage, and put forth leading edge credible research design on the minimum wage—finding no sizable employment effects.
32. Although these examples review modest increases in the minimum wage, we argue that small businesses will likely benefit from the more sizable wage increases that would occur if the FJG were enacted. Businesses would benefit because the program would increase the purchasing power of working-class households, who have a higher marginal propensity to consume. Additionally, with a higher wage, the economy would likely see more robust productivity growth.
33. Adjusted to 2016 U.S. dollars.
34. Additionally, the Kinder Institute ran a poll in the Houston MSA region to assess voter support for government programs that reduce inequality. Seventy-six percent of respondents agreed that “The government should see to it that everyone who wants to work can find a job” (Kinder Institute 2016).
35. A local guarantee as a pilot program would benefit greatly from federal funding but could also be valuable as a standalone poverty and unemployment reduction program by providing immense value added through the dual goals. Extensive data collection and recordkeeping would be essential, allowing rigorous evaluation of the impacts, as well as afford a common starting point for supporters and skeptics alike.