Issues With The Job Guarantee

The Job Guarantee (JG) is a policy whereby the government guarantees public jobs to all people at a decent wage. It is meant to function like an employer of last resort. I have raised concerns with this idea elsewhere, but have never written a post about it. So let me lay out those concerns here.

  1. Jobs Require Capital. It is not the 1800s anymore. You can’t give someone a job by handing them a hoe and sending them out into a field. These days most worthwhile work takes place in the context of complicated production projects that require a good deal of capital (buildings, machines, tools, etc.) to operate. It’s hard for me to see how this capital can be mobilized in a Job Guarantee situation where people just randomly show up to the presumably thousands of JG offices across the country with all sorts of skills and for indefinite periods of time. Some advocates of the JG say that they intend on backing each worker with $50,000 of capital. But this is a dollar figure, not tractors, cranes, buildings, vehicles, etc. What kinds of physical stuff will be at each JG office? Will there just be a capital shed with all sorts of knick knacks sitting around to be mobilized depending on who shows up to the JG office that week?
  2. JG Jobs Must Be Essentially Unnecessary. If the JG is going to act as a employer of last resort, then what that means is people should move on and off the JG and that the size of the workforce on the JG will greatly differ across the business cycle and throughout the country. Because you can’t rely on the JG for any kind of predictable of constant flow of workers, this means that the jobs you have JG workers do must be essentially unnecessary. That is, they must be jobs that we can do with or without. Any job that was very important to do is not the kind of thing you could rely on the JG to fill. Instead, you’d hire someone permanently for it, e.g. with public teachers. Given these dynamics, the JG winds up in a weird limbo position. If the job is an important one that you really need done, then you shouldn’t use the JG to staff it. However, if the job is not so important, then you have a hard time explaining why it should be done at all. The JG thus would seem to only make sense for these middle-ground jobs that are kind of nice if people do them, but really not a big deal if they don’t. I don’t know how many jobs that describes, but it seems kind of narrow.
  3. JG Jobs Must Be Short Term. Related to point (2), because the JG workforce should theoretically turn over a lot and shrink a lot in cyclical upturns, this means that any work that is only valuable over long-term stints is basically ruled out. Training someone up for a job when they may bail at any time doesn’t make much sense, nor does embarking on any project that will fail if people bail in the middle of it. Of course, the possibility that people may bail is a problem all firms face. But the JG does not have the same ability of firms to protect themselves against that possibility. The JG would pay a fixed wage that cannot go up (this is an essential part of its anti-inflation scheme), meaning that, unlike a firm, it could not try to entice the worker to stay by offering a raise. Additionally, because the JG operates off of whoever shows up, it could not, unlike a firm, go out and specifically try to hire someone with the same skills as the person who is bailing. Given this, all jobs must be short-term, low-training type jobs that can provide value relatively immediately.

So, taken together, the JG seems to me like it would consist of low-capital, short-term jobs that are not that important to do. That is, it takes on the sheen of a make-work program that doesn’t even probably build much in the way of skills. Add that to the fact that there will absolutely be times across the thousands of JG offices in the country where there simply isn’t anything to do and you are setting yourself up for some rather unfortunate outcomes.

What’s particularly strange to me about the JG approach is that there seems to be a much more common way to boost public employment without running into these kinds of problems. Instead of guaranteeing an undefined job to whomever, the government would actually put together projects whose purpose it is to hire a lot of people and then go out and specifically hire people for those projects. To the extent that the goal is to hire certain types of people who are especially vulnerable to unemployment (e.g. young high school graduates who aren’t going on to college), you would set the project up with that in mind and screen who you hire to try to bring in the targeted population.

By establishing a set project, you can reasonably determine the necessary capital and mobilize it for that purpose. You can also make sure it is somewhat valuable and worth doing because you don’t have to worry about people dropping out (as you can hire new people). And you can do a longer-term project for the same reason: you are actively going out and hiring people, not just trying to find work for whoever happens to show up. This targeted Public Works approach, which can be ramped up and down cyclically as needed, seems to largely overlap with the JG goal while avoiding many of its seeming pitfalls.

An Actually Meaningful Homogeneity Argument

When you point out that other countries grow as fast (or faster) than the US does and innovate as much (or more) than the US does, all while having higher tax levels, lower poverty, and lower inequality, eventually partisans of laissez-faire capitalism come around to say those countries are all one-off exceptions owing to their homogeneity. Taken as a political argument, this is fair enough: I have little doubt that the existence of a good number of poor Blacks and, more recently, Latinos coupled with a viciously racist society makes it politically difficult to aim for lower inequality. As a policy argument, however, it doesn’t really make sense: more money makes you less poor no matter your skin color and the egalitarian projects less homogeneous countries do try (e.g. public pensions, child benefits) work like a charm.

There is one actually meaningful policy argument when it comes to public policy and homogeneity, however. You can find this argument littered throughout anti-socialist texts during a period where heavy amounts of direct government provision of all sorts of goods was still a more commonly sought-after thing. It goes like this: in a more homogeneous society, direct provision is more viable because people’s specific needs and tastes in goods are more uniform; in a more heterogenous society, such provision will be much less successful because needs and tastes are much more varied.

This is a decent enough argument, but it obviously doesn’t reach cash transfer income, which recipients can spend on whatever they’d like. More interestingly, it seems that those on the side that like to make the vague gestures towards homogeneity are the most likely to favor more direct government provision of goods.

Food stamps are an excellent example of this. My position, and the position of any serious leftist I know, is that food stamps should be converted into a cash benefit program. Not every person with low market income needs more food (they may have other ways of getting that cheaply) and many could use those resources for other important things like paying bills, buying diapers, and so on.

It is generally the conservative, however, who not only balks at the idea of converting food stamps into a cash program, but who actually seems to want to make the food stamp program even more restrictive. It is conservative states that are trying (it seems mainly illegally) to greatly restrict what people can purchase with food stamps. These restrictions bring the benefit much closer to providing a uniform bundle of specific goods, which is precisely the kind of thing that is not supposed to serve a heterogenous society very well. Where everyone in a country basically has the same food culture and eats the same things, a very restrictive bundle of food may work perfectly fine (as it matches everyone’s tastes anyways). Where people in a country come from very different food cultures (some vegetarian, some kosher, some halal, and many derived from the national cuisines of all sorts of other countries), this does not work nearly as well.

Thus, on a policy level, the homogeneity insight is not without its lessons. Finland’s maternity package, which delivers a specific bundle of baby clothes and other essentials to expectant mothers, may not be viable here where preferences in baby clothes probably vary much more, both because of cultural variation and even variations in regional climate. It’d probably be smarter to give expectant mothers a $500 check instead, which incidentally is an option in Finland anyways in lieu of the maternity package. This lesson, then, is not a lesson against egalitarianism in heterogeneous cultures, but a lesson for much more cash-driven programs, which the conservative bloc that raises the homogeneity arguments are also paradoxically against.

A Basic Welfare Framework

Often, it’s said that welfare states are dizzyingly complicated. So many programs! Impossible to rationalize! But this isn’t necessarily true. You can make welfare systems that are bizarrely complicated, but laying out a basic welfare framework doesn’t have to be a byzantine affair. Consider the following four-part framework that captures the vast majority of what a good welfare state seeks to do.

1. Health Care for All
Everyone receives health care regardless of income.

2. Education and Childcare for All
Between infancy and adulthood, every child receives education and childcare regardless of parental income.

3. Income for All
The population is made up of working people (~45%) and nonworking people (~55%). Working people receive market incomes, though sometimes inadequately. Nonworking people do not. Thus, the welfare state should ensure that nearly every nonworking person receives some kind of transfer income.

For children, that’s a per-child cash benefit paid to parents. For elderly people, that’s an old-age pension. For disabled people, that’s disability benefits. For college students, that’s living grants and/or public loans. For jobseekers, that’s unemployment benefits. For caretakers, that’s various caretaker allowances.

Insofar as different transfer income schemes are generally used for different kinds of nonworking populations, this part can seem to be very complicated. But, the basic idea is very simple: attach a transfer income to every nonworking person.

4. Leave for All
Everyone receives paid, job-protected leave for maternity/paternity, sickness, and vacation.

This doesn’t capture all you need a welfare state to do, but it captures the overwhelming majority of it. When you start at the program level, it can look very complicated. But if you start at a sufficient level of abstraction, it’s really not: health care, education and childcare, income, and leave. Add on some means-tested income transfers to pump up the plight of low-income workers and last-ditch public assistance, and you are basically there.