:format(webp)/cdn.vox-cdn.com/uploads/chorus_image/image/59535051/GettyImages_931026840.1524775424.jpg)
Job guarantees: so hot right now.
At least three likely Democratic presidential candidates have now endorsed the idea of offering every American a government job at a living wage. First, Sen. Kirsten Gillibrand (D-NY) endorsed the idea in an interview; then her Senate colleagues Cory Booker (D-NJ) and Bernie Sanders (I-VT) introduced legislation to make it happen. Polling suggests large majorities of Americans like the idea. And in a sure sign that it’s gaining momentum, many libertarian, center-left, and leftist critics are vocally denouncing the proposal.
Some of the disagreements about the plan come down to deep questions of values and goals. When critics like Brian Riedl, the campaign economist for the presidential campaigns of Mitt Romney and Marco Rubio, fret that offering everyone a good job will lead to “pressure to introduce a higher wage or certain benefits that the private sector doesn’t offer,” proponents say: Yeah, it will; that’s exactly the point.
Whether you think that pressure will lead to a better living standard for workers or calamity for private business, and which of those you care more about, is the crux of the disagreement.
But there’s also tremendous disagreement about the idea among people who accept a large government role in the labor market, who think that America’s current economic model is letting down too many workers and leaving wages and labor force participation too low, and who are okay with spending hundreds of billions more a year to help poor and disadvantaged Americans. In some ways, these are the more interesting disputes, among people with fundamentally similar values who have vastly different ideas of what guaranteeing everyone in the country a good job would mean.
Here are a few fundamental questions the debate has surfaced, and what they mean for the proposal.
1) Do we need a job guarantee to reach full employment?
At the core of job guarantee plans is what should be, in theory, an uncontroversial idea: that everyone who wants a job should be able to get one.
Full employment has been a formal legal goal of the US government since the Humphrey-Hawkins Act of 1978, and traditionally, achieving it has been the province of the Federal Reserve. When the Fed prints more money, the result is more money in the hands of businesses and consumers, who spend it and increase demand, which in turn leads businesses to hire more people to make more stuff and provide more services to meet that demand.
A number of leading economists have argued that these tools alone should be enough to achieve full employment and rising wages across the economy. Some have argued that a new approach by the Federal Reserve, such as announcing that it wants to target nominal income rather than inflation or targeting 4 percent inflation per year rather than 2 percent, would go a long way, even without Congress acting.
Others have argued that Congress needs to get involved and the Fed needs new tools to ensure full employment, like, say, the ability to distribute money directly to Americans rather than disseminating it by buying up bonds from banks.
:format(webp):no_upscale()/cdn.vox-cdn.com/uploads/chorus_asset/file/10727223/942761078.jpg.jpg)
This isn’t purely theoretical, either. Sweden, Canada, and Israel escaped the worst of the 2008-’09 financial crisis largely due to good, aggressive central banks. And lenient monetary policy by Alan Greenspan in the late 1990s helped cause one of the few periods of full employment and healthy wage growth in recent American history.
All of which raises the question: If we can get rising wages and jobs for nearly everyone who wants one by changing monetary policy, at no budgetary cost and arguably even without further action from Congress, why spend more than $500 billion a year guaranteeing jobs through a jobs guarantee of the kind proposed by Sanders, Booker, and Gillibrand?
The idea of a job guarantee originally comes from a movement in economics known as Modern Monetary Theory (MMT), one of whose tenets is that monetary policy more or less does not work. Stephanie Kelton, a prominent MMT economist and adviser to Bernie Sanders, has repeatedly argued that the Fed is the wrong agency to task with achieving full employment, partly because it’s not as powerful as it needs to be to do that, and partly because Fed chairs have, in practice, been insufficiently eager to use their powers to help workers.
From that perspective, the answer to the question of “why a job guarantee rather than a better Fed” is obvious: A better Fed can’t do much of anything, in this view. But MMT is a relatively fringe position within academic economics, and most monetary economists do think the Fed can do an awful lot, perhaps even everything necessary, to achieve full employment. If the mainstream is right, shouldn’t we try that approach first?
2) Is a job guarantee the best way to employ people with work barriers?
Even in a world with perfect Fed policy, there’ll still be some unemployment. “Full employment” doesn’t mean literally mean “every single person is employed.” There will still be caregivers, retirees, and severely disabled people who aren’t in the workforce, and due to regular job churn and seasonal work, some working people will be unemployed in any given week. Ski instructors might not work in July, scuba instructors might not work in December, and so forth. When I left the Washington Post, I was “unemployed” for two weeks before I started at Vox because I wanted to visit friends and family for a bit. It was fine.
But there’s some structural unemployment that’s decidedly not fine because it’s not desired by the people who aren’t employed. Plenty of disabled people, for instance, would like to work but face structural barriers, like a lack of work accommodations or an inability to pay for health aides on the job, or a lack of work in a sector where they’re physically able to do the labor. (For instance, if you use a wheelchair and the only work available is as a server at a restaurant or a sales associate at a pharmacy, you might be out of luck.)
Recently incarcerated people face barriers both due to the stigma associated with incarceration and because they’ve been out of the workforce for a while. Some mothers who would like to work lack adequate child care to do so.
Job guarantee advocates argue that the program would help people in these kinds of situations find work and reintegrate themselves into the workforce. And they might — but helping people with significant work barriers is difficult, and might not be best accomplished by a universal program, most of whose enrollees won’t have severe work barriers.
A job guarantee plan released by Mark Paul, William Darity Jr., and Darrick Hamilton for the Center on Budget and Policy Priorities states that enrollees will all have access to a high-quality health insurance plan. But will they also have access to health aides and other support services, in case of disability? Most health insurers that aren’t Medicaid don’t cover those kinds of supports. Will the job guarantee offer free child care for mothers who want to work?
Many job guarantee plans suggest child care and elder care as professions that enrollees could take up. But those jobs require significant training to do effectively, and in any case, a job guarantee is not necessarily the same thing as a child and disability care guarantee.
Moreover, there’s a question here of targeting and cost. Getting work for people facing significant barriers to employment is tough, though we have a few models that can work. But they generally require careful design and effective targeting to the needs of the groups in question.
A few years ago, Georgetown’s Indivar Dutta-Gupta, Kali Grant, Matthew Eckel, and Peter Edelman reviewed the past 40 years’ worth of evidence regarding subsidized employment programs in the US, including programs that (like a job guarantee) employed people directly through the government, as well as programs that got enrollees jobs in the private sector while supplementing their wages and offering services to help them stay working. These programs came in all shapes and sizes but typically targeted specific populations like the formerly incarcerated, people with disabilities, and single mothers.
The report is positive overall on subsidized jobs, and found four models that passed a cost-benefit test: a 1980s demonstration meant to help single mothers train as home health aides, the Center for Employment Opportunities (CEO) in New York that targets ex-prisoners, the New Hope program mostly enrolling single moms in Milwaukee in the 1990s, and a 1980s program to help people with intellectual disabilities find work.
But there was an equal or greater number of other programs with similar goals and, in some ways, similar designs that didn’t prove cost-effective or have any effects on earnings or employment of which to speak.
The Transitional Jobs Reentry Demonstration, which offered subsidized transitional jobs to formerly incarcerated men in Detroit, Chicago, Milwaukee, and St. Paul from 2007 to 2008, briefly increased employment early on but had no effect on long-run employment in traditional private or public sector jobs. Recidivism didn’t fall, and after two years, only about a fifth of people getting help in the program were formally employed (the same share as people in the control group who got support services but not subsidized jobs).
Even CEO in New York, which significantly reduced recidivism, didn’t have significant effects on employment and earnings outside the initial subsidized job itself. It’s a cost-effective program for sure — I don’t mean to knock it, and its recidivism effects are great and important — but it didn’t have the labor market outcomes you’d hope for.
All of which is to say is that designing these programs is really hard: It’s easy to get it wrong and difficult to get it right. That’s not a case against investing in targeted programs to help people with work barriers. The Georgetown evidence review is clear that such programs can work and effective models are worth scaling up.
But the difficulty of implementing them well might be a reason to spend a smaller amount of money scaling up particularly effective programs and rely on the Federal Reserve to ensure full employment and rising wages otherwise, rather than create a universal job guarantee at greater cost that might be less effective at reaching these specific groups, since they’re not its main focus.
3) Will job guarantee jobs be taken seriously and respected?
:format(webp):no_upscale()/cdn.vox-cdn.com/uploads/chorus_asset/file/10727215/885651426.jpg.jpg)
Almost every evaluated subsidized employment program mentioned above, including ones that weren’t cost-effective overall, increased employment while it was subsidized. A fair rejoinder by a job guarantee proponent would be, “We’re proposing subsidizing employment permanently. So why wouldn’t the employment gains be permanent too?”
It’s a fair point. But it also forces us to ask what the goal of a jobs program is. If the goal is to get people stable work, likely to earn them raises and the possibility of a progressive career, then public sector jobs appear to be a uniquely bad approach.
A meta-analysis by labor economists David Card, Jochen Kluve, and Andrea Weber of employment subsidies and other “active labor market programs” in the US, Canada, Europe, and other rich nations concluded that while private sector wage subsidies and training programs can be quite effective in the long run, “Public sector employment subsidies tend to have small or even negative average impacts at all horizons.” In other words: Public sector programs can employ people so long as they’re publicly subsidized, but people aren’t likely to roll over from those positions into durable work elsewhere.
In a recent Twitter thread, Tim Bartik, one of America’s leading labor economists, concurred, writing, “It is probably easier to have rollover into regular jobs from subsidized private jobs.”
Intuitively, it’s easy to see why this might be the case. Suppose a woman is unemployed and goes to a government agency in search of a job since she’s not able to find one in the private sector on her own. The government then goes to CVS and says, “We want you folks to take a chance on this person. We’ll pay a third of her salary if you give her 35 or more hours a week and provide training so she can move into more advanced roles in the company.”
Over time, you can imagine that woman doing well at her CVS job, enough so that CVS keeps employing her without needing government money. She could even go on to get a better job at Walgreen’s without any government intervention because the program has built up her skills and given her experience at a respected firm.
Now suppose that, instead, the government agency gives her a job at Government Pharmacy. Even if she does well there, CVS and Walgreen’s could look at an application from her in the future and say, “Well, she worked at Government Pharmacy, and they will employ literally anyone. So how do we know she’s any good? Why should we take a chance on her when she hasn’t been able to hack it in a private sector job in this field?” That could be a recipe for her to be stuck working at Government Pharmacy indefinitely.
Maybe that’s okay, especially if Government Pharmacy offers good jobs with the prospect of long careers. Many job guarantee advocates envision job guarantee jobs as exactly that: real public sector positions with the possibility of advancement and an upward career trajectory. But it’s also possible that in practice, the work will be stigmatized and viewed by other employers and the rest of society as a kind of dole. That would wind up hurting the workers who sign up, and could deter people who need help from signing up in the future.
Indeed, job guarantee programs bear a certain resemblance to “workfare” programs in the 1990s, where cash welfare was replaced with subsidies given to poor households in exchange for the head of household actively seeking work. Leftists and liberals generally hated those proposals at the time, viewing them as a scheme to take benefits away from people who find it hard to work.
A job guarantee wouldn’t be taking existing benefits away from people; it would be conditioning new benefits on work. That’s a very significant difference, not least because the pay for job guarantee jobs would be much greater than welfare benefits used to be.
But if a job guarantee plan winds up stranding people in $15-an-hour government jobs when they’d be happier either not working or working in jobs with clearer future prospects, then it would somehow combine the worst of normal welfare (whose high phaseout rates trapped recipients in a situation where they couldn’t work on top of it even if they wanted to) and workfare (which imposed strict and punitive requirements on households whether or not they could work).
4) Is a job guarantee the best way to boost worker power?
:format(webp):no_upscale()/cdn.vox-cdn.com/uploads/chorus_asset/file/10727179/630525186.jpg.jpg)
The single most encouraging study on the idea of a job guarantee comes from India, whose National Rural Employment Guarantee Scheme is meant to offer a permanent, guaranteed source of income for rural farmers during the dry season. A group of economists — UC San Diego’s Karthik Muralidharan and Paul Niehaus and the University of Virginia’s Sandip Sukhtankar — conducted a randomized experiment and found that the program raises earnings for low-income households by 13.3 percent, mostly because it bid up wages in the private sector. It also increased employment in the private sector, amazingly.
One reason that could’ve happened is monopsony. That’s a situation where there are only a handful of firms competing for workers. Because of that limited competition, those firms are able to set wages artificially lower. If you’re a lion tamer, and the only circus in town is Ringling Bros., then you’re going to get lower wages than if Ringling Bros., Barnum & Bailey, and Big Apple Circus are all competing to hire the best lion tamers. That’s true even if the Fed is doing its job and the economy is growing fast.
In that scenario, if a job guarantee comes in, it can increase competition for workers and increase both wages and employment in the private sector as different employers keep bidding up wages and hire more people to help make up for highly productive workers they lose to other firms.
A number of recent studies have concluded that monopsony is a very common condition in the American economy. That suggests that a job guarantee could have a similar effect in America as it did in India, raising wages across the board.
Sounds good! But given the difficulty of designing an effective jobs program, it’s worth wondering if other policies can do the same thing. We’ve known since at least the 1990s that in situations of monopsony, increasing the minimum wage can boost wages without reducing employment, a result that Econ 101 would say is impossible. For workers higher up the income chain, encouraging unions and collective bargaining and establishing wage boards to set standards for specific industries can serve the same purpose, boosting worker power against companies and limiting firms’ power to set wages.
Finally, good antitrust enforcement can help fight monopsony. The market for lion tamers took a hit when Ringling Bros. and Barnum & Bailey merged in the early 20th century. More pertinently, hospital mergers can hurt wages for nurses and medical assistants by denying them the option of quitting and going elsewhere. An antitrust regime that fought such mergers aggressively could wind up boosting wages.
Most job guarantee advocates would endorse these policies on top of a jobs guarantee. But given that they’re all costless to the federal Treasury, and many can be implemented by states alone even with the federal government in Republican hands, it’s worth asking how much of the benefits of a job guarantee can be achieved by cost-free regulation rather than a huge, hard-to-implement program.
