Deficits don’t matter, spending does

Bringing the Great Deficits Debate down to Earth.



One of the most interesting debates (to nerds like me) in the #EconTwitter-adjacent discourse concerns the importance of deficits. In case you’ve been living under a rock, it goes like this: deficit hawks say that the federal government continuing to spend way more than it brings in (by borrowing money) is reckless and bad. “Modern Monetary Theory” (MMT) proponents say that deficits are basically fine as long as you print your own currency as the U.S. does; the only real danger of excessive spending is inflation, and inflation can be managed by taxation which takes money out of the system.

After a long while thinking about it, only recently have I formed even a tentative opinion on the matter.

Here’s why the hawks might be right. Say the U.S. runs a $1 billion deficit by promising to pay some mix of random people and world governments $1.1 billion in ten years. That is, quite obviously, $1.1 billion dollars that the next generation can’t use for themselves.

Even hawks concede that deficits are sometimes ok. If you have, like, a war to fight against Nazis, then it probably makes sense to borrow some money and ensure your survival. Nonetheless, every dollar we borrow now is a dollar that our children can’t use for themselves. Even more, it’s a negative-sum trade since we have to pay interest on top of the principal we borrow.

I hope I’m not bastardizing the MMTers’ argument here, but I think they’d say that our children and our children’s’ children and so on can run deficits themselves, ad infinitum or until AI takes over and kills us all. If some generation goes crazy and borrows a quadrillion dollars to finance fully automated luxury space communism, the gates of hell don’t open up and swallow the Treasury. Instead, we run into the actual material productive capacity limits of the economy and everything gets more expensive as people compete for peoples’ time and stuff: inflation. And, if this happens, we just tax people to take money out of the system until the CPI chills out a little bit.

Back to the Basics

One of the gripes I have with economics as a discipline is its tendency to explain things only in terms of abstract concepts like inflation and debt. These are useful tools to be sure, but at the end of the day these things reflect actual resources being deployed. To understand deficits, then, we have to think about what they represent in the real world. Let’s try it out.

Suppose the federal government spends $150 billion but only brings in $100 billion in taxation and other revenue so we borrow the difference. On the ground, this means that the government is deploying $150 billion worth of resources—paper, fighter jets, VA nurses, whatever. Where do these resources come from? How is it that the government can buy stuff with money it doesn’t have?

For simplicity, let’s pretend that those $150 billion can be broken down into $50 billion each of paper, jets, and nurses. And, let’s assume, tax revenue buys the paper and the jets whereas deficit spending/borrowing pays for the nurses.

I’m no fan of the libertarian adage “taxes are theft,” but in a demoralized sense they kinda are. Basically, the government just takes the paper and the jets from people. Yes, in reality people voluntarily hand over tax dollars because social contract and We Live in a Society and then the government buys stuff from consenting firms and individuals. But, functionally, this is just “taking” with a few extra steps and political philosophy papers in between.

What about the nurses, then? Well, the government doesn’t want to just appropriate/kidnap/tax the nurses into existence, so they borrow the nurses. Well, in real life they borrow money from hedge funds and Asian governments and American citizens and then pay the nurses, but in this stylized model we can pretend that they borrow the nurses from themselves. In other words, the government tells the nurses “if you work at our hospitals this year, we’ll pay you next year+interest” and voila, we’ve spent $150 billion while only taking/taxing $100 billion.

Now, let’s take things back to the level of economic abstractions. The hawks say “we have to pay the nurses next year, with interest, and interest is getting to be 100% of GDP!” and the MMTers say “anything we can do, we can afford to do!” and “next year we can borrow money again.” They’re both right, but the MMTers have a key insight. “Anything we can do we can afford to do” suggests a corollary: “if we can’t afford to do it, then we can’t do it.”

Deficits aren’t magic

Deficits can’t make real resources materialize out of thin air. We have to pay the nurses next year, yes, but the nurses have to exist first for us to “borrow” them. If we didn’t “borrow” them, they’d be doing something else like watching TV or working at a private hospital. So is deficit spending good or bad? It depends to some extent on the interest rate, sure, but first and foremost it depends on whether we as a society would rather the nurses work for the VA or do whatever else it is that they’d be doing.

At one level, this is incredibly banal. Of course it matters what the government spends money on. But when the economic discourse deals in abstract concepts like “deficits” far removed from things happening in spacetime on planet Earth, this point can get lost in the weeds.

Why MMT is 70% right

According to Wikipedia,

Economist John T. Harvey explained several of the premises of MMT and their policy implications in March 2019:

  • The private sector treats labor as a cost to be minimized, so it cannot be expected to achieve full employment without government creating jobs, too, such as through a job guarantee.

  • The public sector's deficit is the private sector's surplus and vice-versa, by accounting identity, which increased private sector debt during the Clinton-era budget surpluses.

  • Creating money activates idle resources, mainly labor. Not doing so is immoral.

  • Demand can be insensitive to interest rate changes, so a key mainstream assumption, that lower interest rates lead to higher demand, is questionable.

  • There is a "free lunch" in creating money to fund government expenditure to achieve full employment. Unemployment is a burden; full employment is not.

  • Creating money alone does not cause inflation; spending it when the economy is at full employment can.

I put what I think to be the key points here in bold.

Deficits can make dollars and cents materialize out of thin air, but they can’t make VA nurses or fighter jets do the same. At the end of the day, government spending and deficits are social technologies designed to organize economic production into one configuration instead of another. And MMT is right about this.

If the whole country is lounging around and feeling sorry for itself and the government borrows money to start paying people to do useful things, this is 100% a “free lunch.” Until we reach full employment, there are “idle resources” that society would be better off using, which the government can use deficit spending as a tool to activate.

Even at full employment, deficits might be a good method to rearrange how resources are being deployed. Climate change is the obvious example. If nobody is investing in carbon capture and clean energy research (public goods with positive externalities), the government should tax, borrow and spend in order to get actual scientists in the real world to do this kind of research, deficits be damned.

Of course, rearrangement isn’t free. That’s what interest on the debt reflects. But we can’t forget that interest payments themselves are a transfer of resources from one entity (the government) to another (bondholders), not an exhaustion or destruction of them. Remember, the government could, at least theoretically, recuperate these losses through taxation.

Why MMT is 30% wrong, kinda

  • Creating money activates idle resources, mainly labor. Not doing so is immoral.

  • There is a "free lunch" in creating money to fund government expenditure to achieve full employment. Unemployment is a burden; full employment is not.

I know this is a talking point of degrowthers and anticapitalists, but economic output really isn’t the ultimate arbiter of social value. Right now, I’m enjoying myself and learning quite a lot by writing a blog post that nobody will pay for, and hopefully you readers are doing the exact same. But, since no money changes hands, none of this value “counts” toward GDP.

Now, it is totally reasonable to retort that unpaid voluntary labor and leisure aren’t “idle resources,” but let’s think about this for a second. Depending on the job, there is absolutely a wage at which I’d close my laptop and start doing whatever the government was paying me to do. MMT, I think, would call that a “free lunch” even though it’s at least on the dollar menu.

Maybe I’m being unfair to MMT. Just as important as “no economic value doesn’t imply no real value,” though (sorry for the triple negative), is the point that “economic value doesn’t imply value.”

Suppose I were instead sitting inside scrolling through instagram, an activity that doesn’t help others and that I neither enjoy nor dislike. And, all of a sudden, the government borrows some money from China and starts paying me $20 an hour to dig ditches and fill them back in again. I find the task unpleasant, but it’s worth the $20 to me.

True, I’m better off because of this arrangement, but society as a whole (which includes me) is worse off. Why? Because the only difference between this and the status quo ante is that one person (me) is now doing an unpleasant activity instead of a neutral one, and $20 worth of resources have been transferred from someone else to myself. The only thing those $20 are buying is toil.

Perhaps this point is buried in an MMT textbook somewhere, but it seems to me that the whole MMT paradigm would support this “Aaron digging ditches” arrangement. Before the deficit spending, I was an unemployed, willing and able worker. After the deficit spending, I am being paid for my labor. There’s no inflation because we’re not bumping up against the productive capacity limits of the economy, so what’s the issue here?

Deficit or no deficit, the issue is that society as a whole has been made worse off. I think this illustrates my main point: deficits don’t matter, spending does. And by “spending,” I mean spending resources. Deficits aren’t inherently anything; they’re a good social technology when they rearrange real life resources in a good way, and a bad one when they rearrange real life resources in a bad way.

Is this obvious?

Perhaps this is abundantly obvious to all the economists I think I’m critiquing. I honestly wouldn’t be shocked if this were the case. Regardless, I don’t think the message is getting through to the general public. If a nerd like me who reads Noahpinion blog posts and listens to Macro Musings in his free time doesn’t find it at all obvious what the Great Deficits Debate is really about, who exactly does?

The upshot

“Ok”, you might say, “you’ve used terms like ‘real life resources’ and ‘the real world’ about a dozen times now. What is the real life, real world policy takeaway? Good question, and here is my attempt at an answer: the potential creation of a deficit is an argument neither for nor against the enactment of some policy. The answer should be the same irrespective of whether the Treasury happens to have enough money sitting around to pay for it without borrowing.

Take aducanumab, the controversial and super expensive Alzheimer’s drug I wrote about recently. Although the FDA partially walked back its approval, over a million Americans (mostly on Medicare) will be eligible for the medication that probably doesn’t work. Should medicare pay for the drug or not?

Although most people would probably say it depends greatly on the drug’s efficacy, deficit hawks would probably be very wary of adding another few billion to the deficit. MMTers, on the other hand, are more likely to call the decision a no-brainer; after all, is anyone harmed if we borrow some money to pay for the drug?

Though I hate to play Enlightened Centrist, the real question here is whether the development, production, and distribution of aducanumab is a good use of real people’s time, glass vials, medical school hours, syringes, and the thousands of other resources that go into pharmaceutical production. If they are, deficit spending would rearrange our complex economy into a better configuration. If not, it is worse than useless.

Even if we’re not at full employment and paying for the drug doesn’t threaten inflation, it is at least possible that society as a whole would be better off lounging around and scrolling through TikTok than getting its shit together to make a fancy drug. After all, people like scrolling through TikTok, even if it doesn’t add to the economy. And every minute that med school professors are lecturing and glass vial producers are producing glass vials is a minute that someone, somewhere isn’t doing something she likes to do. Deficits don’t matter, spending does.

Abstracting up, abstracting down

Chemists zoom out from subatomic particles to discuss the behavior of molecules. Biologists zoom out from molecules to discuss proteins, tissues, and organisms. A few more layers up, we get sociologists zooming out from individual behaviors to reach the realm of “social capital” and macroeconomists zooming out from individual transactions to discuss “GDP” and “deficits” and the like.

This “zooming out,” or “abstraction” as it is sometimes called, is important, but it has to be a two-way street. Just as it can be useful to talk about “baking soda” and “vinegar” instead of their subatomic components, it can be useful to talk about “deficit spending” instead of ground-level resource deployment.

Nonetheless, moving “down” the chain of abstraction from molecules to atoms or from macroeconomic terms to specific supply chain and labor market descriptions is not just an important robustness check (if your explanation of baking soda and vinegar’s reaction is forbidden by fundamental physics, somebody is wrong) but also an intellectually generative process.

This is understood well enough in the physical sciences and even at the psychology-neuroscience nexus, but perhaps insufficiently in the realm of (macro?)economics and sociology. Explaining something in terms of “GDP,” an “accounting identities” and the “price level” might be useful, but it isn’t sufficient. At best, it leaves out an interesting piece of the puzzle. At worst, it can lead to bad policy that causes suffering in the real world.

I would certainly be keen to see people smarter than myself dissect high-level economic and sociological phenomena into their ground-level mechanisms to see if they still make sense. Although Modern Monetary Theory isn’t perfect, I give its proponents credit for nudging economics a bit in this direction.


Our Puritan heritage has imbued "self control” with a kind of moral halo, but I think this has led us astray. Modern Monetary Theory sounds analogous to a snake oil panacea or get rich quick scheme - something too good to be true, and “fiscal discipline” sounds like something a wise, battle-hardened elder would espouse.

But some things that seem too good to be true are not. Vaccines are “too good to be true.” The cost-effectiveness of charities like Against Malaria Foundation (save a life in expectation for a few thousand bucks) is too good to be true. In fact, the last three centuries of economic growth is too good to be true.

For better or worse, the United States Federal Government has the power to print its own currency and shift around money in any way it likes. Like deficits themselves, this power is inherently neither good nor bad. No, we can’t pretend interest doesn’t exist, but neither should we delude ourselves into thinking that transferring resource control to bondholders is equivalent to setting paper and fighter jets on fire, or paying someone to dig holes and fill them back in again.

The fruits to be gained from improved social coordination are tremendous. Just look at that graph! There are a lot of things wrong with the modern economy, but the US government’s ability to borrow and spend is one of our most powerful ways of picking that fruit. Let’s not throw it away.

Remember: deficits don’t matter, spending (resources) does.

Leave a comment