As America celebrates yet another Labor Day, it’s hard not to dwell on the fact that many in the country are still looking for work. According to April figures from the Bureau of Labor Statistics, 9.8 million people actively looking were unable to find employment. Creating jobs for them might be the easiest possible way to help the millions of impoverished job seekers out of poverty.
So, what if we just hired everybody who needs a job? Could a federal job guarantee work?
But portioning out blame among 50 years of economic and policy changes is impossible. Who knows where today’s poor would be if it weren’t for the way globalization, technology and de-unionization have marginalized low-wage jobs? Government anti-poverty programs may have made a situation that would have gotten even worse a little better, even if politicians like Ryan claim the size of poverty programs proves these programs foster an ambiguously defined feeling of dependence. In the extreme, Ryan’s rhetoric is a way of blaming the victim: Low wages create conditions that make it very difficult for parents to finish their education or for fathers to invest in their children’s lives, and then Republicans say this lack of education or parental involvement causes poverty.
Paul Ryan and others on the right blame poverty on a lack of responsibility from the nation’s poor; the left generally blames poverty on a lack of money. A nationwide jobs program offers a way of directly quantifying what increased spending we’d need to reduce unemployment, and indirectly assessing how culpable the poor are for their predicament.
Once we start investigating the numbers, it’s surprising how doable guaranteeing jobs looks. If there are 9.8 million unemployed people, giving each of them a $30,000-a-year full-time job (approximately $15 per hour) would cost about $294 billion. On the one hand, that’s a lot of money. It’s approximately equal to the $265 billion the federal government spent on Medicaid last year, and it’s about half the $585 billion spent on Medicare. These are large federal programs, and squeezing an extra program of their magnitude onto the federal books would not be that easy. On the other hand, the federal government took in almost 10 times that amount in total revenue. It’s only about 1.8 percent of total GDP.
These are imprecise estimates for a number of reasons. First, jobs cost money, since we have to provide inputs to workers before they begin. More important, we don’t have a good measure of how many people would sign up for a guaranteed job, since there are many who are marginally attached to the labor force and thus fall through the BLS’s measures. These people might come back to the labor force were a job guarantee in place. Even then, a large jobs program doesn’t have to be all-encompassing. We could deliberately create a $50- or $100-billion jobs program to bring down unemployment to healthier levels. The same analysis applies, just with lesser magnitudes.
A jobs guarantee would save a lot of money, through reduced utilization of other social benefits programs. Since existing programs exclude many of the poor and unemployed, though, calculating how much we’d save is tricky. About 2.6 million people receive unemployment benefits, less than one-third of the total unemployed. Of the unemployed, about one-third are long-term unemployed, meaning they’re likely ineligible for unemployment benefits. A good guess might be that 1.8 million of the unemployed in the Labor Department’s number are eligible. Since the average unemployment benefit is around $300 per week, giving these people a guaranteed job would save about $28 billion a year.
Guaranteed jobs would also reduce the Medicaid and food stamp rolls. Since food stamp eligibility never ends, most families receiving guaranteed jobs would eliminate someone from the food stamp rolls. A typical three-person family (a single person with two kids) collecting food stamps received a $400 benefit in 2011, so 9.8 million guaranteed jobs might reduce food stamp spending by billions. That same three-person family costs an average of $10,000 per year to insure via Medicaid, but since eligibility criteria vary widely between states, not every family receiving a guaranteed job would subtract from the Medicaid rolls. (If, for example, we assume about half of them would, 9.8 million guaranteed jobs subtracts another $49 billion from Medicaid costs).
In sum then, guaranteed jobs could potentially pay for a third of their costs through reductions to safety net costs. There would also be increases to government revenue through taxes. If new jobholders pay 8.5 percent in payroll taxes, such taxes would raise another $25 billion. They’d additionally pay some relatively small amount of their income in income taxes and contribute to GDP in ways that would increase taxes indirectly.
To find more savings from guaranteed jobs, we have to look hopefully to ways a solid income impacts social behavior. A 2004 paper by Jeffrey R. Kling, Jens Ludwig and Lawrence F. Katz found that poverty in neighborhoods significantly increases the likelihood of young people in that neighborhood engaging in criminal activity. Since it can cost over a quarter-million dollars per year to incarcerate a teenager, even small reductions in criminal behavior can result in tremendous cost savings.
But the more important offset to the costs of guaranteed jobs would be the output the new jobs would produce. What could guaranteed jobs do? For starters, they can provide all sorts of benefits the market fails to provide. For an example we might look to the Civilian Conservation Corps, a New Deal program that constructed buildings for national and public parks, built rural roadways, and led nationwide reforestation efforts. Such projects today could include urban renewal projects, since urban blight leads to higher crime rates.
The conservative objection to jobs programs is usually that such win-win scenarios are impossible. Writing in National Review, Kevin Williamson writes that what guaranteed jobs provide, “by definition, is: nothing of economic value.” But this objection only holds if the market perfectly provides all human wants at the correct human prices. Consider the case of cleaning a park versus cleaning a house. If somebody pays someone else to clean her house, we unquestioningly include such activity as economic activity and include the transaction in our national accounts, and conclude the purchaser must get some benefit to a clean house. A community almost certainly gets a similar benefit when a public park is cleaned or a dilapidated building is rehabilitated. The difference is that since no one person owns the park, no one person realizes the economic benefits to cleaning it, and thus no one pays to have it cleaned. A government job fixes the market failure of non-coordination.
For a more real-life example, consider an unemployed person in an inner city. He wants a job, but will revert to stealing food if he can’t get one. His neighborhood is full of dilapidated buildings, and if they are cleaned up, eventually developers will build new properties in the area, revitalizing this neighborhood. The developers, however, don’t individually have enough incentive to start the renewal process. The guaranteed job thus serves as a conduit to getting conditions that are beneficial to all parties. The worker gets a job fixing blighted properties, the community gets better building opportunities down the line, and the government saves on prison costs. The economic activity that is eventually generated could easily be greater than the cost of the job.
Even when the output is not worth the cost of providing the job, the government recoups some of the money through the cost savings described above. This “invisible subsidy” should rationalize all sorts of jobs the private sector won’t. And the fact there’s any output makes a guaranteed job usually superior to other government benefits. The conservative case against social benefits is that they represent a deadweight loss, paying people not to do work they might otherwise supply to the marketplace. Making work itself the benefit blunts this problem, adding at least some output to the economy and giving workers an incentive for each hour of labor they provide.
There’s reason to believe even conservative presidents know the value of supporting the job market through government employment. We can compare the Reagan recovery that began in November 1982 and the Obama recovery that began in May 2009. Five years later, the Reagan economy had seen an increase of 8.3 percent in government employment; the Obama economy has seen a decrease of 3.3 percent in government employment. If Obama’s government had followed Reagan’s path, there would be an additional 2.6 million employed people by now, enough to lower the unemployment rate below 5 percent.
We can look abroad and to the past to see real-life ways the government can run jobs programs. After its 2001 economic meltdown, Argentina created a jobs program called Heads of Households, which paid the unemployed to participate in community service, maintenance and jobs-training programs. Heads of Households cost 1 percent of GDP but provided jobs to over 5 percent of the population, and lowered national rates of extreme poverty. During the New Deal era, the United States created jobs programs like the Works Progress Administration and Civilian Conservation Corps. These programs constructed thousands of new schools, libraries and city halls, many of which are still in use today. We could use such a program now to construct much-needed new schools or nursing homes.
A jobs program today might look something like an expansion of AmeriCorps, which coordinates thousands of community development projects across the country. They focus on the areas of environmental protection (including pollution cleanup and installation of recycling facilities in poor neighborhoods), health (driving patients to doctors’ appointments or health screening for schoolchildren), and economic opportunity (building low-income housing, mentoring or financial counseling). These projects counteract some of the negative ripple effects of poverty described above. Since the economic downturn, interest in AmeriCorps has grown tremendously, so instituting a larger jobs program may be as simple as increasing its small $1 billion budget and hiring more people.
Of course, the government already has a large program that’s treated as a jobs program: the military. Republicans who pooh-pooh stimulus as wasteful spending are among the first to defend the need for the military to create jobs. “Cutting our military,” according to Republican congressman Buck McKeon, “brings grave economic costs,” because of the effect on employment. A top rationale for joining the military is a lack of employment opportunities elsewhere. But once we acknowledge the military is partially a jobs program, we have to acknowledge it’s an expensive and unproductive one. It would be much cheaper to hire people to do something like community service, and we’d invest the money at home rather than on foreign wars and foreign military bases.
How is this any different from most stimulus programs, which usually cite employment as their primary motivating factor? Most important, popularity and efficiency. The stimulus was very unpopular, probably owing to how vague and indirect its effects were, but a job guarantee, according to a recent poll, is favored by a majority of those expressing an opinion. This means advocating for such a guarantee would be something of a home game were Democrats to adopt it into their platform. And it could directly target the desired economic outcome of greater jobs. The purpose of the 2009 stimulus was to create jobs via immediate increased demand, but much of the spending would only generate very high-skill jobs (like for medical and energy research) or would take a long time to implement (like Race to the Top education reform). Given the unpopularity of the stimulus and greater popularity of a job guarantee, it’s unfortunate the president took the indirect approach. Instead of a policy of creating jobs by increasing demand, we could implement a policy of creating jobs by creating jobs.
Stimulus is usually understood to be temporary, since most economic theory holds that recessions and shortfalls in demand are temporary. The sluggishness of the recovery these past five years should make us at least question this assumption. Economists Larry Summers and Paul Krugman have argued we may be following Japan into a period of “secular stagnation,” or a permanent shortfall in demand, caused by demographic factors and resulting in a diminishment of consumers and investment. The anthropologist David Graeber caused a stir by noting that many of the employed seem not to be doing any real work but adding to a pointless bureaucracy – a phenomenon he called “bullshit jobs.” Indeed, it’s interesting to note that the supposedly fake jobs created by the stimulus (like infrastructure repair and renewable energy installation) seem much more like real jobs than the supposedly real jobs that make the most money (like creating credit default swaps and doing estate tax planning). If these factors are real and borne out by the next half-decade of experience, we may have to accept that market downturns and a lack of work are not always temporary and indefinite fiscal support should be the new normal.
What might Republicans get out of passing a jobs program? An ability to test their beliefs. Paul Ryan discusses poverty in terms of dependence and a lack of familial commitment. If the poor are truly unwilling or mentally unfit to work, such a program would fail. Workers would walk out of their guaranteed jobs or not even sign up, and Fox News would be there to report it. Republicans might also support the notion of having people work for their government benefits rather than receive their government benefits. Of course, if a job guarantee is a wild success, Republicans might have to reassess their prior belief that unemployment is a failure of workers — and not a failure of the market.