Breaking down MMT’s Guaranteed Jobs Scheme for Fighting Inflation

For “Ask an Economist” this week, I have a question from John O. about Modern Monetary Theory (MMT). I’ve purposefully avoided MMT in the past for a few reasons. First, because it appears to be under-theorized. In other words, often when economists try to talk with MMT advocates, the advocates argue by retreating to previously unstated assumptions.

Second (relatedly), as economist Per Bylund put it: “In scholarly research, interest in MMT is limited… One reason for this is likely that MMT focuses on policy prescriptions rather than explanations… which makes it unsuitable for research.” This makes conversation about it difficult.

Regardless, I wanted to do my best to answer John’s inquiry. He says:

MMT holds that a Federally-funded, locally-administered Job Guarantee is a way to achieve full employment (everyone willing and able to work can have a job—not 96.3%, but 100%) and would also act as a price anchor, tending to dampen inflation, just as the buffer stock of unemployed workers does today, but without all the attendant miseries (crime, addiction, suicide, etc.) of unemployment.

So can we have a free lunch of 100% employment and no inflation? No. Let’s see why.

Full Employment and MMT

One of the golden calves of MMT is the idea of full employment—or 0% unemployment. As John notes, this doesn’t mean everyone has a job in MMT World (either in theory or practice). Zero unemployment just means everyone who reports himself as looking for work finds a job (instantly in this case).

The idea that we should have full employment all the time is a revolt against the idea that there will be some natural rate of unemployment due to:

1) people’s skills not matching currently available jobs, or

2) because of short-term changes of employment due to labor market frictions.

The idea of the natural rate of unemployment is that even healthy economies have people who are looking for jobs for the above two reasons.

MMT advocates deny that this is necessary. Unlike economists, they see these unemployed workers as a wasted opportunity. In the words of MMT scholar Pavlina Tcherneva: “Unemployment holds a ‘special’ status in economic theory, compared to other social deprivations. Economists do not speak of the natural rate of hunger, or the natural rate of illiteracy, or the natural rate of homelessness.”

Except, to be clear, Tcherneva is essentially wrong about this. While it’s true we don’t measure things like the natural rate of illiteracy or homelessness, economists generally recognize that bringing the rate of crime, illiteracy, or homelessness down to zero is prohibitively costly.

Page through any introductory economics textbook and you’ll likely find a quip that “the efficient amount of anything is not zero.” That includes crime. It also includes unemployment. Unemployment is not unique in that sense like the quote above implies.

Certainly we prefer people to find jobs that suit them sooner, but sometimes the cost of finding a job immediately is higher than the benefits.

MMT advocates claim to fix this with a simple solution: a government-paid jobs program. Many iterations of such a program have been proposed, but their common feature is that the government-created jobs pay enough money that unemployed workers actually want them. Sometimes a “living wage” is proposed, but we should recognize that a living wage wouldn’t be enough to entice, say, an engineer between jobs.

How would the government pay for this? Ultimately, in MMT, the answer is that the government can print money to pay. We’ll come to issues with this later.

So MMT pursues full employment as a goal and claims the guaranteed jobs program will do so while “anchoring” inflation. Why? The idea is that if people are spending a lot (because they all have good-paying jobs) there will be lots of employment and therefore no need for the government-guaranteed jobs. As a result, spending from the government will decrease (and therefore money printing will as well).

This is the sense in which MMT advocates claim the guaranteed jobs policy is checking inflation. Spending increases by the public are offset by the spending decrease by the government, so prices stay stable.

This may sound good on the surface, but it breaks down as we look a little deeper.

Employed for What?

There are two issues with this argument. First, the full employment described by MMT is not a good goal. Second, it cannot reliably temper inflation. These problems are related.

To see the issues with this system, consider what it would mean if we implemented it today. The government creates a new jobs program. Anyone who wants a good-paying job can get one and be paid in freshly printed US dollars. The new jobs created must exceed to jobs eliminated by higher taxes, otherwise we wouldn’t achieve full employment.

What happens? New money enters the pockets of these workers, and they go and spend it. As they spend it, you have more money chasing the same amount of goods, meaning that you are going to have higher inflation.

One of the classic lines of MMT is that printing money does not cause inflation unless there is full employment. Our little thought experiment here shows that moving from some unemployment to full employment does indeed involve inflation. Historical precedent of events like stagflation (where the economy experienced increasing monetarily driven inflation and high unemployment) also shows that the MMT line isn’t true.

Why do MMTers say this then? Well, the logic is that as you hire people to these new government jobs, they start producing more as well. Inflation is caused by more money chasing the same number of goods. But what if we have more money chasing more goods produced by the new government jobs?

In other words, MMTers claim that while this program does increase the supply of money, that increase is offset by an increase in the supply of goods. These two cancel out, leading to no inflation. Our problem is solved. There’s no inflation because the new production of goods and services makes up for the new money-printing, and everybody has a good-paying job. With this policy everyone will get rich—and quick!

Sounds nice—doesn’t work. Here’s the problem. The government can create jobs, but jobs are not in and of themselves a good thing.

Here’s an example. Imagine the new guaranteed jobs program pays government workers to dig a hole and fill it in all day every day.

Did employment increase? Sure. Did the supply of goods and services increase? I guess if you consider a newly dug and filled-in hole a good, then yes. The problem is that this isn’t really a product that consumers want. It doesn’t make anyone better off.

If you included this good in your inflation quantification, its price would certainly fall. But since no one wants to buy the unfilled and refilled holes, it really shouldn’t be included in the inflation quantification. Since the new production was wasteful, it doesn’t lower the prices of the goods relevant to consumers. As such, inflation does not actually fall, even if some quantifications which included newly dug and refilled holes indicate that it did fall.

In other words, this new production isn’t enough to curb the inflation that matters to real people. You need the right production. This oversight is a microcosm of a larger issue with MMTers where they ignore the territory of a healthy economy in favor of the map of its quantifications.

At this point it would be fair to point out that MMTers likely have slightly better proposals for what sort of jobs to create than a program that digs holes and fills them back in. Maybe we could have the new workers make hamburgers, cars, solar panels, or some tangible goods. Would that fix the issue?

No. The whole problem is that creating jobs is an engineering problem, as the government can get more output (jobs) by simply putting in more inputs (money); creating jobs which aren’t wasteful, however, is an economic problem.

Of course, if the government spends more money, it can make more jobs. But which jobs are best?

Private, for-profit business has a mechanism to solve the economic problem of which jobs are best. If a private business hires a worker whose work output is not valued sufficiently by consumers, the business will make a loss on that worker. This loss signals to the business that the job does not create sufficient value to consumers to merit the costs associated with the job (such as wages and the worker’s use of other inputs).

A government-guaranteed job, no matter how it is facilitated, does not have access to the profit and loss mechanism.

As such, there is no way to tell if the output from the job is valued sufficiently by consumers to justify the loss of that worker in another line of production. Producing hamburgers will probably create a good that consumers like more than dug and refilled holes, but the point is that insofar as consumers were…