How Modern Monetary Theory changed my mind

Johnny Bowman
17 min readSep 5, 2018

Modern Monetary Theory (MMT) completely changed my mind about how the US government can dramatically increase equality and the overall wealth of Americans. To every “socialist” policy out there, it answers the “how will you pay for it?” question. What’s more, MMT’s answer is rooted in monetary operations — an approach I personally find more revolutionary than hunting Mark Zuckerberg down and extracting his tears. It turns out we don’t need his tears, or his profits, to pay for things. They’re just nice-to-haves.

The downside to MMT’s logic is that once you grasp the basics it seems so obvious that those who disagree seem a little slow. Like they’ve done 1/5th of a homework assignment and called it a day. I wish I was more sympathetic, but I’m an insufferable MMT enthusiast myself, so my only hope is to make you an insufferable enthusiast too. Explaining MMT is a little dry though so before I go into that here’s some context around why it’s a thing these days.

How MMT proved it was kind of a big deal

For any theory involving money to prove its mettle, it’s gotta do 1 of 2 things in 2018.

  1. Make somebody a truck load of money
  2. Predict a crisis

MMT did both. And if you’re stiffening your skeptical spine right now I’ve got you right where I want you so good. Read on.

MMT is a set of interrelated legal and macroeconomic ideas that started getting cohesive in the early 90s. Warren Mosler developed a lot of this cohesion himself while running his own hedge fund and made $100 million in the early 90s betting that Italy wouldn’t default on its own debt, because they could print as much Lira as they wanted. For the 80% of folks who read that and thought $100 million is a lot of money, yes you are correct. For the 20% who read that and thought $100 million is nothing to write home about, first of all, we’re talking 1992 money here, which is damn solid. Second of all, you’re a pig, and you know it.

On the crisis front, MMT economists predicted the Euro crisis like nobody else did. By the early 2000s, MMT economists had predicted that European countries sharing the same currency and the same rules would make it harder for member countries to dig themselves out of a recession. More specifically, they wouldn’t be able to spend their way out of the recession like the US did and Germany’s labor/austerity policies would make it harder for other EU members to increase wages, prolonging the crisis. This is exactly what happened, and you can read about it here.

And for a theory to really cement itself in the annals of monetary history it helps if its origin story involves Donald Rumsfeld in a steam room, which probably makes a lot of MMT aficionados uncomfortable, but I’m here for it. As for why it’s gotten popular in the last year, the recent history of MMT’s ascent is better explained here.

Why the MMT approach is different

The singular frame that helps explain MMT thinking for me is accounting. As MMT economist Scott Fullwiler puts it, “every transaction in a real-world economy affects financial statements of those engaged, and if an economic theory or a posited model is not consistent with how real-world financial statements are affected, then the theory is inapplicable.” Reread that sentence (or not, fine) because it’s a little dense but the message feels like it should have been a big deal in like 1800, not in modern times. Scott’s arguing that money leaves a paper trail everywhere it goes. Every party that transacts in money will see their financial statements affected by those transactions, dollar for dollar. So if you want to figure out how money works, stop your theorizing and follow the paper trail. It’s bonkers that this argument was novel when Scott wrote it in 2012 and that today 99% of economics departments in higher ed do not teach this as an approach to understanding financial causation.

Here’s how MMTers put this kind of thinking into practice. In this paper by MMT economist Stephanie Kelton, she walks through how government spending affects accounts at the Fed, the Treasury, and commercial banks. By looking at which accounts at which institutions go up and down over time, she concludes that taxes and taking out loans do not actually fund the US government. That’s a direct refutation of the macroeconomics 101 class I took at Harvard Business School. Counterintuitively, the US government creates the money first, then spends the money, and then uses taxes and bonds to drain excess reserves from the banking system to hit our target interest and inflation rates. Even if the phrase “uses taxes and bonds to drain excess reserves” has you hitting the snooze button, her first point is a gamechanger: the US government doesn’t need taxes to spend money. I’ll explain that below, but taking a moment here for this emoji: 🤯

Ok so what does MMT actually mean?

For a theory built on the technicalities of government accounting, the major takeaways are not that technical. I’ll dig into the core ideas that personally make me excited for it, and how these ideas differ from previous thinking. I like to group these ideas by the major questions it answers:

What is the limit of sustainable government spending?

The answer I learned in business school was the deficit. Seems reasonable. If spending is greater than tax revenue, we have a deficit, which is paid for by borrowing from other countries. And while this allows us to pay for things in the short term, our kids and their kids will be stuck paying back the US’ creditors, hampering the US’ ability to spend and invest at a later day. In other words, deficit spending is sacrificing our future for present comforts.

Under MMT, the public deficit is not inherently bad — it is in fact necessary. It sees spending in general as a balance sheet, comprised of the public, private, and foreign sectors. A deficit in one must be offset by a surplus in another. Here’s how these deficits and surpluses normally add up:

  • The private sector is in surplus because households and businesses wish to save for the future.
  • The public sector runs a deficit — we’ve done so every year since 1789 with only 7 short exceptions.
  • At least since the Reagan years, the foreign sector runs at a surplus, otherwise known as a trade deficit (or current account deficit for the economists).

The academic phrase for this view of the economy as a group of sectors that balance each other out is “sectoral balances”.

As economist L. Randall Wray sums it here, “if we want our private sector to save, which almost everybody agrees is a good idea, the public sector must run a deficit.” And what happens when the public sector runs a surplus? The US has done so 7 times, and Wray has an answer: “The first 6 of those were followed by our only 6 depressions.” And the 7th was during the Clinton years, which led to a recession in the early 2000s and then a global financial collapse in 2008.

So if public deficits are necessary to avoid a depression in the private sector, can the US government spend unlimited amounts of money? No. Because inflation is indeed a real limit. If national spending vastly outpaces our ability to nationally produce things, the public sector competes with the private sector for spending on resources, which raises prices — also known as inflation. If a buncha inflation happens without folks’ incomes rising, people get angry.

The ideal level of government spending under MMT is the amount necessary to buy up the economic output of the United States at full capacity — any more and we get inflation — but any less and we’re leaving material wealth on the table, which usually disproportionately affects our poorest citizens.

And just to be clear, deficits can coincide with inflation, but do not necessarily cause inflation. A good example of this is Japan. Japan’s total debt/GDP ratio is 253% and their deficit/GDP ratio is 4.5% — the highest in the developed world. The US’ total debt/GDP ratio is 105% and our deficit/GDP ratio is 3.4%. If deficits caused inflation, Japan should be experiencing inflation big time, but Japan’s most recent published inflation rate is 0.7% — even below their 2% inflation target. The US’ inflation is 2.9%. Clearly, the deficit is not a great predictor of inflation.

When are we at risk of rising inflation?

Both MMT thinkers and current Fed leadership agree that when spending power exceeds our full production capacity, inflation likely rises. However, they differ on how to measure when we’ve reached “full” production capacity.

The Fed’s current policy measures production capacity through unemployment rates. Currently, full production capacity is benchmarked at a “full employment” rate of 4.5% unemployment over the long term. I used to think this made sense because it’s also referred to as “the natural rate of unemployment” and why mess with it if it’s natural? As the Fed’s theory goes, if unemployment dips below that, we’re at risk for higher inflation, because presumably this is the point employers start thinking they need to raise wages to hire the candidates they want. The Fed uses this unemployment target to control a target inflation rate, which is currently 2%. When unemployment gets “too low”, the Fed makes it more expensive for commercial banks to borrow money, which means loans to businesses and households also get more expensive, which makes it harder for households to buy things and businesses to spend money and hire more people.

MMT objects to this method for two reasons:

  1. 4.5% unemployment is a poor proxy for full production capacity.
  2. There’s nothing “natural” about 4.5% unemployment. It’s a policy decision that’s needlessly cruel as it uses our most marginalized citizens — the unemployed — as a “buffer stock” for controlling inflation.

On point 1, it’s MMT thinkers’ perspective that full employment should actually means 100% of people who want to work, are working. The unemployment stats obscure underemployment and people who would otherwise work, but have given up, and so increased spending at 4.5% unemployment would not necessarily lead to wage gains. We can see this in employment stats and wage data. If you look at the percent of people who actually have a job, as opposed to those who qualify as “unemployed,” we are nowhere near our pre-recessionary level of employment. Regarding wage data, while we’ve done a fabulous job of controlling inflation since the 1970s and keeping “unemployment” low, real wages haven’t budged. For 40 years, despite hitting “full production capacity,” employers have not raised wages except to keep up with inflation.

On point 2, it worries MMT advocates that we’ve chosen permanent unemployment for a set percentage of people as our primary means for controlling inflation. Other buffer stocks exist, and MMT argues that a job guarantee would be much better. Instead of people going in and out of employment with business cycles, people would go in and out of publicly guaranteed jobs. In bad times, more public jobs. In good times, people leave those for private jobs. What’s more, this would be a lot faster than the trickle down effects of our current strategy of lowering interest rates. In a recession you could get a public job rather quickly as opposed to waiting years for the economy to pickup again. For the US economy to be strong, having a “stabilizer” that works swiftly is important — otherwise people literally get depressed and sick, exacting huge mental, physical, and economic damages.

One final note on inflation. Unless you remember World War II — the last time our economy was really humming at full production capacity — inflation during our lifetimes has had little to do with US employment. Often times prices rise for important staples (like energy and housing) due to external factors. For example, in the late 1970s, inflation reached 14% . This was due to the Iranian Revolution causing oil prices to spike. In 2011, despite unemployment still being high, inflation in the US bumped up a bit and Ben Bernanke himself was quick to explain it was due to commodity spikes, not excess government spending. Today, housing, healthcare, and education prices continue to go up for reasons completely outside of employment patterns. In these modern cases, unemployment is even less effective as a buffer stock to control inflation.

Don’t increasing deficits lead to greater default risk?

Oh boy. Caring about this is the ultimate test of considering yourself a responsible business person (of which I am one). We should absolutely not default on our debt — I can’t think of a way for our Treasury to simultaneously screw more people than defaulting on our debt. This is why Howard Schultz, Starbucks CEO and prospective Presidential candidate in 2020, claims our national debt is our biggest weakness. He’s a responsible business man and we have a lot of debt. I mean he also says that because he believes he can run for president, fix it, and cement his legacy as America’s billionaire savior, but he’s still pretty worried. Fortunately for America, and unfortunately for Howard’s ego, it’s impossible for us to default.

Lemme elaborate as this is when MMT really starts departing from mainstream thinking. There’s a strong condition to not defaulting, which is that the nation in question has to be financially sovereign. Financial sovereignty means all the following are true:

  1. You own your own currency (e.g. the US has dollars, the Japanese have yen)
  2. Your currency is not pegged or tied to anything else like gold or another currency. It “floats.”
  3. You owe no debt in currencies besides your own (e.g. the US’ debt is all in US dollars).

In other words, if all our liabilities are in US dollars, and we can “print” money at will (“print” is a loaded term in MMT since what actually happens is digitally crediting and debiting accounts at the Treasury and Fed), then we can’t go bankrupt. Why would we? We will always create money to pay back our debt. As long as we are financially sovereign, we are at zero risk of default unless Congress pulls a Ted Cruz and stops paying people for no good reason.

Now, if we spend too much, we do get inflation. Inflation is real and is currently screwing Venezuela left, right, and center. You get inflation when you spend way more than your country’s productive capacity is capable of matching. Like if there were 100 computers in an economy and $100 of spending money, every computer would be $1. If there were 100 computers and $1 million of spending money, every computer would be $10,000. Creating more money doesn’t make people richer if there’s not more stuff to buy.

If deficits don’t matter, what’s the purpose of taxes?

The status quo conception of taxes is that the government takes in money, usually via taxes, in order to spend it. Under MMT, the government spends money into existence, then taxes. National output, not taxes, fund the government. The underlying theory here is chartalism and has had various proponents over the last 100 years (such as this one by a former Chairman of the New York Federal Reserve, who also is the father of income tax withholding). Chartalism states that the point of taxes is not revenue, but for the state to direct economic output in a certain direction.

A famous example of this is colonial Brits trying to get an African village to mine metals for them. One way to do this is to hold people at gunpoint and demand they start mining. Classic strategy, but this requires a lot of soldiers. What the British did instead was create a hut tax, payable in British pounds. How could villagers get British pounds? By working in the mines. And if they didn’t, death and destruction followed. This (real life) story changes the narrative for why we tax in two ways: 1) the British needed to pay wages in pounds before they were able to tax in pounds and 2) taxes were what created demand for British pounds. Without taxes, it is tough to convince people to use a currency —the failure of most coins created by ICOs is an example of this — although Bitcoin is proving taxes aren’t always necessary to drive currency adoption.

Taxes, in other words, may not fund the US government, but are a very useful (and potentially exploitative) tool. In the modern economy, taxes do a few things: curb inflation (taxes lower demand for goods), disincentivize certain behaviors (such as a cigarette tax), and redistribute wealth. This is true whether you subscribe to MMT or not, but MMT thinkers don’t just believe this is a byproduct of taxes - they believe this is the whole point of taxes. As a result, our tax policy should be designed to optimize for these social outcomes. Similarly, government spending should be optimized for full employment, not to balance the budget. This view that we should design taxes and government spending around explicit social outcomes stems from functional finance, a theory that builds on chartalism and was developed by Abba Lerner in the 1940s.

There are big policy implications of this frame. For one, we don’t need taxpayers to run the US government, we need natural resources and people who make goods and services. Mitt Romney famously complained that 50% of America doesn’t pay taxes (which was wrong because 82% of us pay taxes due to payroll taxes but roll with me here). The implication was that 50% of the US is paying for the goods and services of everyone else. Under MMT, the notion of “taxpayers’ money” is false — public money is created by the US government with zero help from tax payers. Taxes exist to generate demand for the currency and achieve certain social outcomes. It would be more accurate to say that taxpayers have money that leads to inflation and inequality, two things we want to check, and so they’re taxed. This explanation is dramatically unsatisfying to anyone who’s ever paid taxes, but it means that the richest among us are not paying for the poorest among us. Raul Carrillo and Jesse Myerson expound on this and the identity and class implications here.

If we can spend more, what should we spend it on? A job guarantee? Universal basic income?

Policymakers could use MMT to justify spending money on any number of government sponsored programs, including free college, Medicare for All, free childcare, universal basic income etc. And many MMT economists do support some or all of them. But if there’s one idea that promises to reduce inequality the most within the MMT framework, it is the job guarantee. Here are the main reasons, contrasted with the other popular anti-poverty idea making waves these days — universal basic income (UBI).

  1. Higher wages. If the limit to government spending is inflation, and increasing national output lowers inflation, then we can spend more if we create more. A guaranteed wage under a federally funded job guarantee could therefore be much higher than a federally funded UBI. Tcherneva provides a mathematical model for why that’s the case here. This paper from Wray, Dantas, Fullwiler, Tcherneva, and Kelton models what the program would actually look like and finds that if the job guarantee employed 15 million people and paid $15 / hour for full time work (~$30,000 a year), plus benefits, inflation would only rise 0.74 percent. Meanwhile, supporters of UBI generally support UBI incomes that are much lower, due to inflation. Andy Stern’s UBI proposal and Sam Altman’s UBI field study, which are more generous than most UBI proposals, propose $12,000 per year. UBI as a result is strictly an anti-poverty measure, while a job guarantee could get folks closer to a middle class lifestyle.
  2. Better health, social mobility, and education. Less incarceration and suicide. Unemployment causes serious negative secondary effects. These include higher suicide rates, increased sickness/healthcare costs, declining mental health, increased incarceration rates, and lower social mobility and educational attainment for children of unemployed people. Studies showing the results of these effects are explored in Tcherneva’s paper here. All these secondary effects carry social costs that get worse during economic downturns, and so employment can be seen as a social good. And while we don’t have enough data to make claims about UBI’s effects on these social outcomes, the bar is pretty gosh darn high.
  3. Better help for our most marginalized communities. While UBI would certainly help reduce poverty and improve social outcomes among all demographics, its inherent focus is everyone — not communities who have been held back the most. Simply put, when it comes to addressing the barriers for black wealth accumulation, most of our policies — whether they be education, skills, or family focused — rest on assumed deficiencies of black people themselves, not the ongoing structural barriers that prevent black wealth accumulation. While a job guarantee doesn’t fix racism, it provides a direct link to jobs as opposed to relying on education or the private sector — which can have their own racist barriers. In a paper published by the Federal Reserve Bank of St. Louis Review, economists Darrick Hamilton and William “Sandy” Darity Jr. walk through the myths of the Black wealth and employment gap and are joined by Alan Aja and Daniel Bustillo in explaining why these gaps are best addressed through a jobs guarantee here. And of course Black communities are not the only ones who stand to gain. Aja, Carrillo, and Rita Sandoval explain why Latinx communities should care about the job guarantee here. For a by-the-numbers breakdown, The Wray, Dantas, Fullwiler, Tcherneva, and Kelton paper estimates who would participate in a job guarantee, broken down by race and gender. They find that Black and Latinx communities and women would directly benefit the most out of a job guarantee plan.
  4. A better stabilizer when the economy tanks. Private markets go up and down in cycles, creating a greater need to address unemployment and poverty in the down times, and less of a need during the good. UBI does nothing to counteract these cycles, while a job guarantee is at its peak usefulness during the lowest moments of our economy.
  5. Socially useful projects. The private sector is not capable of creating goods and services that people want when it’s difficult to capture value (aka get people to pay for it). Cleaning the environment is one big example, and solutions are generally government-sponsored (e.g. Pakistan just pledged to plant 10 billion trees in 5 years to fight climate change). Our own health is another. To be clear though, the “what will people actually do” under a job guarantee is still a hotly debated issue and is the question I’m personally most interested in these days. And to all the Pessimistic Patricks out there saying there just aren’t enough jobs that the government could provide to employ folks, the options are as limitless as human needs are. All these jobs need to be is useful for this policy to work. There most definitely are enough needs and jobs out there to fulfill a job guarantee.
  6. It competes with bad jobs and doesn’t subsidize good ones. If McDonalds had to compete with a job guarantee, what would they do? Either raise wages or automate. Both are totally acceptable outcomes with a job guarantee in place. If there was a UBI, there would be a higher chance McDonalds would decrease wages than increase wages. Given UBI without a job guarantee, any pressure to raise wages to a living wage standard would be due to federal or state laws making them.

To reiterate, the job guarantee is not mutually exclusive with UBI — many MMTers want both given some people are not able to work and yet have every right to live above poverty. However, there is a general belief that a job guarantee would provide for a higher standard of living for the poorest among us, would create a more inclusive economy across race, gender, income, and age, would make recessions suck less, and could accomplish some truly useful goods and services that the private sector is not willing to do.

And if any of this makes you want to talk/learn more about Modern Monetary Theory a) let’s get in touch. I’m johnnybowman at gmail dot com b) the 2nd International Conference of Modern Monetary Theory is September 28–30 in New York City. Please come and I’ll see you there.