Modern Monetary Theory (opens new window) proposes a few principles that seem pretty undeniable to me:

• Unemployment is wasteful. Society loses the opportunity to benefit from the labor of many of its members, and doubly must contend morally and otherwise with the social damage that comes from the deteriorating skills/employability (opens new window), poverty, distress, illness, and resentment of the unemployed. Further, "natural" unemployment of the able-bodied is a myth, especially since current monetary policy enforces unemployment as an inflation control measure (opens new window).
• Money is a social technology, not a scarce resource. Money is only valuable because taxation exists (opens new window), representing our collective decision to direct some portion of our economic activity toward projects that create public goods. Money's value directly corresponds to the amount of labor and resources (leveraged to greater power with technology and infrastructure) available in the nation that creates it, not from some other entity such as gold or markets. The amount of money held by any actor merely represents their share of the "promises" for labor and resources in the underlying society, or in other words their purchasing power (opens new window).
• Deficits or "inflating" the currency can be productive. If an expenditure of money (or in other words an application of labor and resources) in some way increases the productive capacity of the underlying economy, then "wealth" has been created and so extra money also needs to be created to represent that new wealth. So "inflationary" creations of new money won't actually cause actual inflation if the money is used to build infrastructure or some other productive form of leverage.

And outside of the world of MMT, economists of many stripes have long known that public goods are in general underprovisioned (opens new window) (there's a lot of useful public goods work to be done that isn't being done).

However, one of the cornerstone policy proposals of the MMT crowd that addresses these realizations is a Job Guarantee (opens new window), which is hard to explain quickly but here's my best try:

• Government guarantees every member of society a job at some wage/benefit that is low but livable, intended as a social floor (opens new window).
• We decide what these jobs are (and avoid make-work (opens new window)) using a distributed system where nonprofits and government agencies send job descriptions, and some process is used (that's hopefully democratic?) to approve those jobs if they would produce some public benefit. I have to admit to being somewhat unclear on the exact mechanisms here.
• Unemployed people can apply for and be directed to any of these approved jobs as they need. The intention is for these jobs to ramp up during economic downturns to "catch" the newly unemployed and direct them toward publicly beneficial work. This is intended to be counter-cyclical (opens new window), so as the economy recovers people leave the Job Guarantee system and get other jobs.

I think the goals of a Job Guarantee are good (achieve full employment and increase counter-cyclical investment in public goods), but I don't think the specific system is a good idea. The best phrasing I've found comes from this article (opens new window):

... [T]he idea that thousands of administrations across the country will be able to usefully employ random flows of labor with random sets of skills in random durations is fairly implausible.

In short, meaningful and useful public work or projects can't really be achieved with labor intended to be temporary, with unpredictable skills and experience.

Also, there's a reasonable worry the presence of these low-paying but abundant jobs would make existing public and nonprofit workers precarious, since those agencies would always have some incentive to use Job Guarantee workers instead of market-rate ones.

Luckily, I think we can achieve the goals of full employment and counter-cyclical investment in public goods with two separate but complementary programs, Universal Paid Training and a Public Goods Guarantee. These systems attack the two different aspects of the unemployment/public goods problem in ways that more realistically consider the realities of skills and economic leverage.

Universal Paid Training

Universal Paid Training is simple:

• Any time someone wants, not only when they become durably unemployed, they can apply to train in a particular skill. We pay them the same "social floor" wage they would have received in the Job Guarantee, so they can live and support their family etc.
• This training is done at normal public institutions such as high schools, trade schools, community colleges, and universities, or even perhaps in apprenticeship programs that have been validated in some way (open question). Only using public institutions would be essential to allow us to control the costs (opens new window) and quality of the training.
• To prevent overcrowding and oversaturation of any particular skill, the number of people per time period the program will pay for is limited by comparing the cost of the training to the expected income increase the applicant would receive compared to their existing skills, and adjusting for the expected wage decrease implied by the eventual entry of the current number of promised trainees. This means the program is less willing to pay for smaller increases in skills, for more common bundles of skills, in more crowded or soon-to-be crowded sectors. The cost is time prorated, so all other things being equal, training that takes less time is preferred. Overall this should have the effect of more heavily prioritizing the most useful next marginal skills increases.
• People still have to be accepted to the public institution/program in the normal way, given the capacity of those programs, so this prevents people from being paid to pursue programs they aren't ready to succeed in.

This program isn't intended to only or even mostly be for "college" degrees, at least not in the way we usually think of them. I'm convinced there's been a bit of a "college bubble" (opens new window) in the US for the last 30-40 years. Partially this is due to a cultural shift as physical work lost social status, and as the internet age disrupted sector after sector. But it feels likely to me that if we properly supported the acquisition of physical skills, and then properly invested in many of the currently neglected public goods that would necessarily employ those physical workers (to install universal broadband, guarantee healthcare, update infrastructure, recapture advanced manufacturing, etc), the cultural shift would be overwhelmed by the "facts on the ground".

If we committed to Universal Paid Training, we would do well to dramatically reform education. I already referred to "bundles of skills" intentionally, since I'm convinced we should no longer think about training in terms of monolithic "degrees", but in terms of skill trees (opens new window). This makes it possible to much more granularly understand what skills are actually needed in society. I also think inverted classrooms (opens new window) and more asynchronous teaching are promising, especially since we would want these education systems to be able to accept people all year round.

Public Goods Guarantee

It's so important that I'll restate it: public goods are enormously underprovisioned! (opens new window) It's obvious how this harms society in the form of simple lost opportunity for greater shared flourishing, but it seems clear it harms workers directly in the labor market as well. Since we don't properly coordinate around public goods work, all the people and skills that would be fully leveraged by that work are much less appreciated and in demand. A trite example is the fact that open source software is often built by overworked and underappreciated volunteers (opens new window), despite its immense social benefit. Many other disciplines and communities suffer the same fate.

The Public Goods Guarantee system intends to solve exactly this problem, in a way that's dynamically balanced by the actual cost/difficulty of a project compared against the amount of public desire for it. This ensures the skills generated by Universal Paid Training are reliably put to highest and best use. Specifically:

• Nonprofits and government agencies at all levels can submit project proposals for work they argue would create positive public externalities (opens new window), meaning their benefits aren't being captured privately. The proposals must estimate their costs in terms of jobs defined by the bundles of skills discussed above, and standardized descriptions of materials and equipment.
• Citizens can use some set of democratic weights to support or oppose these projects. These weights are meant to represent the degree of public desire and support for each particular project, and this is used to prioritize them. Specifically the projects are ordered by the ratio of the expected cost against the democratic weight, meaning the projects with the greatest social return on investment (opens new window) are given priority.
• There is a "stream" of funding set aside in the budget that fills a "pool" for these projects. The projects are committed to and allocated funding in order of priority. In every time period, the top number of projects that can be funded with the current "pool" are committed to.
• If projects run out of money they are not automatically granted more, they must submit another separate project proposal. This increases pressure to be realistic, and to build incrementalism and risk mitigation into the project plans (opens new window).
• The "stream" of funding flowing into the "pool" is proportional to the level of unemployment, with some minimum to ensure we're always eventually funding the most desired projects. This creates the counter-cyclical nature of the program, and ensures it is related to the amount of "slack" in the economy. We can afford anything we can actually do (opens new window), but the opposite is also true: we can't afford things we don't have the available labor to do! Specifically the fluctuating amount is set to be enough money that all the current unemployed people could be hired at some slight discount to their current rolling average income, plus some added extra to account for the material and equipment costs that would be implied to hire them.

The nastiest problem for this system is the fact that estimating project costs is notoriously difficult (opens new window), no matter whether the project is public or private, big or small, known or unknown. But people have studied this (opens new window), and there are ways to discipline forecasts to make them more realistic: by doing reference class forecasting (opens new window), so basing estimates on actual similar projects; building incrementalism and modularity into the plan, so difficulties are encountered and overcome in small doses early on; planning slowly to foresee problems before they are difficult to fix; and simply acknowledging the difficulty by adding fixed buffers to estimates.

The Public Goods Guarantee partially enforces reference class forecasting and careful planning by requiring proposals to state their estimates in terms of standardized commodities. It encourages incrementalism and risk mitigating modularity by capping funding and requiring further proposals for more. And the mere presence of democratic transparency and prioritization shines light on proposals and allows the distributed knowledge of the crowd to discipline unrealistic ones.

Since project proposals can be any size, it is possible for the full spectrum of risk vs reward to be encompassed. If a large project can be broken into useful phases, it can start with small proposals and increase ambition as it gains trust and removes risk.

An important part of a Public Goods Guarantee is that it is automatic and democratic. It is possible for Congress or other officials to become gridlocked or otherwise unable to act to invest in public goods even when doing so would be productive. A Public Goods Guarantee ensures productive projects are funded regardless of the health of other public decision-making processes.

These systems are vastly preferable to a Job Guarantee.

Any economic system must keep itself attached to the hard reality of welfare, resources, and leverage. If it doesn't, it can lose the ability to actually create value for people, such as how Lysenkoism denied basic genetic and agricultural science, causing much of the famine deaths in Soviet Russia and China (opens new window).

People are repelled by the idea of make-work, not just out of a notion of fairness of contribution (opens new window) or meaning (opens new window), but because they intuitively understand it could lead to societal disaster. Encouraging large numbers of people to perform useless labor could create a situation where useful work is "crowded out", and society's real capacity to achieve things collapses. People are wary of any system that even seems possible to fall into those dysfunctions, as demonstrated by the continued inability for a Job Guarantee to gain much respect or political momentum.

In contrast, when people gain skills other people find valuable for achieving goals (both private and public), the overall "latent capacity" of the economy increases, because the amount of actual welfare (opens new window) we can achieve with a unit of labor and resources increases. This means Universal Paid Training expenditures not only prevent a harm (unemployment), but also produce an added benefit (long-term progress and societal empowerment).

It isn't an accident education is one of the first and main public goods funded by taxation: it is intended to ensure the children of the poor are not doomed to perpetually remain poor (opens new window). Universal Paid Training simply follows this realization to its logical conclusion, intending to prevent all vicious cycles of poverty (opens new window) that are nonetheless "market efficient" (opens new window). Education and training are the main ways we uplift people sustainably. Knowledge and skills ensure people can contribute to society, and that they have some capability under their direct control (opens new window) to maintain their independence.

Further, these systems would much more realistically "catch" the unemployed in economic downturns. The Public Goods Guarantee scales up the amount of public work when the amount of private work decreases, meaning people are "caught" without even necessarily decreasing their income or status. And for all those who can't find useful employment in either public or private work, they can get the training needed to change that. Every able person left out of the current labor market, regardless of their situation, could be caught usefully by Universal Paid Training, since an inherent required capacity of an educational system is to evaluate the current abilities and possible paths for people with arbitrary backgrounds. Normal programs like old age pensions and disability then cover the remaining people who can't reasonably work.

Both of these solutions place fairness and usefulness of work as a higher priority than guaranteeing full employment. I'm convinced there is a ton of extremely useful public goods work to be done in the world (opens new window), and we could easily employ every able member of society in either doing that work or training to be able to do it. As time goes on the marginal usefulness of the projects being democratically approved would decrease, which is fine. We could then either increase our societal ambition, or decrease the length of the work week or other changes to take more collective leisure. An alternate system (like a Job Guarantee) that less accurately prioritized useful work would be less able to actually transform society in a way where prosperity increases without labor hours increasing.

Further, these systems would make the labor market "automation proof". By directly supporting people to fully reskill whenever the need arises, we retain the benefits of creative destruction (opens new window) without basically any of its enduring harms. I'm very skeptical LLMs or "AI" will replace many workers in any close timeframe (opens new window), but automation in general has long been a social concern (opens new window). We absolutely need a way to protect and support workers that doesn't allow them to unilaterally block actual positive progress (opens new window).

All of these arguments combine to make Universal Paid Training and Public Goods Guarantee both more likely to be effective, and more politically pragmatic as well.

A better world is possible. We just have a lot of work to do.

---

FAQ

This doesn't solve [insert other problem].

Yep! These are discrete system proposals that target particular problems. If you think some other system subsumes these ones, solving the problems I've targeted as well as others, I'm interested in that. But for other problems there need to be other solutions.

Are you proposing any existing social programs be removed if we instituted Universal Paid Training and a Public Goods Guarantee?

No. I already pointed out that old-age pensions (social security) and disability would still be necessary since those people can't meaningfully work. And certainly normal unemployment insurance would still be useful to buffer against transient job loss. Universal Paid Training is concerned with durable job loss, either due to long-lasting downturns in the economy or permanent changes in society or technology.

However it might be necessary or useful to adjust the parameters of disability. Someone might be able to train for different work even if their disability mas made them unable to do what they used to.

And our definition of a "disabled person" might change over time, if automation truly advances to the point where very few narrow skills are still economically useful. It might be a bit clunky, but if we created a notion of "disability due to permanent unemployability" and expanded it over time, we could smoothly scale all the way to a post-work world of "Fully Automated Luxury Communism" (opens new window). The continued investment in public goods would enable us to harness increased automation for the benefit of the public.

But regardless, a lot of our social programs have been massively defunded anyway! We'd be doing strictly better with these systems.

What about "unskilled labor?" Would no one be left to do all of that?

The guarantee being made in this program isn't for training, but for public goods. This system isn't like the Job Guarantee in that every person is guaranteed any particular situation, but merely that the amount of spending on public goods projects is guaranteed to be enough to employ everyone. People are still required to seek and gain that employment, possibly by submitting and leading one of those publicly funded projects!

This means there will be some who can't get into Paid Training (due to its constrained and competitive nature), and therefore might remain "unskilled". As long as we set the minimum wage to the same level of the Paid Training wage, those people will still be fine if they can find any employment (which there will be plenty of because of the Public Goods Guarantee).

And finally, let's remember that "skilled laborers" still often must do "unskilled labor" in the course of their skilled jobs. A backhoe operator sometimes has to "get out and shovel" to handle tricky situations. A software engineer sometimes needs to do tedious data entry or classification. Etc.

What about "independent contractor" loopholes?

They're loopholes and should be closed! It shouldn't be possible for companies to treat people as independent contractors if they have control over their schedules etc. This has already been discussed at length in other places and I don't think it's cogent to this discussion.

Would a bunch of people go to school forever to avoid working?

No, because the training institutions will remain competitive and constrained, and because the system isn't always willing to pay for a specific person/skill pairing based on the calculation of the needs of those skills. Once someone has gained sufficiently employable skills, the program would be unlikely to accept them to continue to gain only marginal increases.

The dynamic and self-balancing nature of this program really makes it unlikely someone will end up in a situation where they both can't get into any training program or find any job, at least without them being legitimately disabled. If rapid changes in society make skills that were previously useful no longer useful, then the system will decrease the current expected value of those skills and therefore increase the expected marginal gain for training in other ones.

Do you really think this is realistic when our existing government is so dysfunctional?

We do have lots of dire existing problems that will require dramatic structural change (new constitution anyone???). I do think those problems are higher priority.

But there's still some hope here. I actually think Universal Paid Training and a Public Goods Guarantee could be much more bipartisan than one would first imagine.

• Economic populism is becoming more popular and within the Overton window, even among right-wing voters.
• There's a strong argument for Universal Paid Training as a source of "national power". There's a case to be made that the huge amount of government investment in science, technology, and public education (GI bill) that came as part of the New Deal and World War II were responsible for creating the golden age of American dominance.
• Universal Paid Training would be massively beneficial to any industrial policy attempting to reclaim advanced manufacturing from China for national security reasons. Conservative thinkers are aware the US can't quickly regain an industrial base because of our lack of trained manufacturing engineers (opens new window) for example.
• Although not "market-driven" in any particular way, the Public Goods Guarantee is at least very respectful of subsidiarity (opens new window), which conservatives claim to care about.

The most harmful toxin we must purge to make our society reasonable again is market absolutism (opens new window), and I'm not going to achieve that in this essay!

Markets are often irrational, do you really think these systems should be market guided?

I am not a market absolutist (opens new window). I fully recognize the many failures markets can and regularly fall into (opens new window), and the ways in which they often structurally lock in early advantages (opens new window). However, it is nonetheless true that any economic/social system must build in some idea of "preference feedback" (opens new window), meaning it has to find a way to determine what people really want and who can actually give it to them (opens new window) that isn't manipulable by pretending to care about things you really don't (opens new window).

Monetary markets, for all their flaws, fit this description, at least if we build the structure of the law and government to fix their main problems, mainly asymmetry of information, externalities, and monopoly power (opens new window). By using (p)redistributive (opens new window) systems like Universal Paid Training and a Public Goods Guarantee, Pigouvian taxation (opens new window) (taxes on negative externalities like pollution), and monopoly control (perhaps even using wealth taxes? I'll have more to say about this in the future), markets could achieve the one thing they're actually good at, accurately signalling who actually wants what and who can achieve it for them.

My assertion is that Universal Paid Training and a Public Goods Guarantee would solve the main problems with the labor market, because it would prevent the most common "races to the bottom" and vicious cycles of poverty.

What centralized agencies or bureaus would be needed to administer these systems?

Universal Paid Training would require increased workload and processes at the Bureau of Labor Statistics (opens new window) to measure the current expected income for different bundles of skills and forecast the impact of more entrants to those skills. I'm confident that work would be much more legible and less manipulable than would be required to process all the Job Guarantee applications.

The Public Goods Guarantee is blessedly decentralized and transparent already, and would only require some small amount of machinery to collect and display project proposals and the democratic weightings over them. A new small bureau could do this work. It would be perhaps useful for some body to evaluate proposals for their realism and rigor, but this would largely be advisory since the democratic prioritization is hoped to provide most of that function. Otherwise it would only really need to be continually watched by Inspectors General (opens new window) to prevent directly fraudulent projects and expenditures.

It seems you're ignoring the arts, would Universal Paid Training make low income professions disappear?

I don't think so. People like and want art of many kinds, and many kinds of art are definitely public goods and have been treated as such in the past (opens new window). In fact Public Goods Guarantee would almost certainly increase the amount of artistic work being done in the country, since the system would fund public goods people want.

The only danger to the arts is that people would just want other types of things so much more than art that it would be neglected and de-prioritized. But if that happened, we would have to simply acknowledge it as a harsh truth and wait for more foundational needs (opens new window) to be satisfied first.

Besides, people have studied the arts on their own dime for basically all of history! I don't think it will be going anywhere. Artists are often "economically irrational" (opens new window), but that is probably to our societal benefit.

What about people who genuinely don't want to work?

The most interesting criticism of these systems is to compare them to workfare (opens new window). There are myriad criticisms of workfare, most of which don't apply here (it isn't only work-focused since it invests in education, it isn't market absolutist since it scales up public goods spending, it isn't "totalising" (opens new window) since beneficiaries remain in society, it assumes the continued presence of worker rights). The one I'm most interested in is simply the concept that any system which requires people to work or face destitution is coercive.

I'm sympathetic to this worry, but it becomes less defensible if we consider the question through the frame of a social contract that bears close resemblance to many used in the past (opens new window):

• Our group/society has agreed to democratically work together to achieved shared aims.
• We demand that in order to benefit from the resources we jointly create, one must contribute to our shared efforts.
• If someone doesn't wish to contribute, they don't have to, but they also can't take resources from us.

I don't think this social contract is unreasonable. Without a demand of mutual obligation, there's a "loophole" in a social contract that can possibly be exploited endlessly. And the presence of a true democracy (which we don't have, read my other posts) ensures that the members can continually tweak the rules to ensure demands and rights are reasonable. In a large society whose members can't reasonably be held accountable by immediate social relations, we need systems of rules instead.

The ability to "opt out" is key, and stands in stark contrast to literal forced labor such as in colonial slavery (opens new window) or the coercive workhouses (opens new window), where people weren't allowed to simply leave. If emigration from a country is always allowed, and as long as some kind of "Zomia" (opens new window) continues to exist, people can indeed opt out of society if they truly prefer a marginal existence to following the rules and demands of any society.

Of course I could be wrong. Maybe human sociality is more complex than a simple "contract", or maybe the primary reason people work is to gain status and so we don't need to threaten people with ostracism. I'm not sure.

But to be clear, what I've proposed seems like exactly the kind of empowering source of equal opportunity that has always been denied to the poor and working class. I have a hard time imagining that most of the current homeless people in the world wouldn't immediately be lifted up by either these programs or better disability support.