Matt Yglesias Is Wrong About Defense Spending & GDP

I don’t usually mind Twitter personality and defenestrated Vox Media founder Matt Yglesias that much. But a couple weeks back he engaged in an argument I’ve been dying to make a video about for years. So with this video I took the opportunity to complain about the irritating data point that Defense Spending has fallen as a percent of GDP, and how meaningless it is in the context of the US federal budget. I feel like I got outside my wheel house a bit, maybe diving a little too deeply into accounting I don’t fully understand, but I think it makes for a fun video.

If you’d like to earn my undying gratitude, please click here to support this project through Patreon. Please do reach out to us through Twitter, Facebook, Youtube, or our e-mail newsletter.

Video Transcript after the jump…

Are you ever just minding your business, complaining about the amount of money we’re wasting on Defense spending, when some know it all pushes himself into the conversation with something that seems smart but is really just incredibly ignorant?

Matt Yglesias usually doesn’t bother me much. I am a 44 years old politics nerd, so I have been aware of his inexplicable micro celebrity for over two decades now. Ever since the blogging days of yore!

We were both wrong about Iraq, but 20 years ago we were pretty firmly opposed ideologically. As we’ve gotten older he’s moved right and I have moved left. I have realized that Reaganism has run out of gas, and Yglesias, as he has left behind establishment outlets like the Vox media empire he helped found, has realized that young progressives can be pretty annoying. So we’re now fairly close on a lot of domestic politics, and perhaps more importantly how sick we are of most mainstream political discourse.

I don’t agree with Yglesias all the time, but I frequently find his substack useful and I often enjoy his endless Twitter provocations. But I get irritated when he points his trademark snark at a topic I care about. We spend waaaay too much on weapons in this country, and a week or two back Yglesias decided to trot out one of the Pentagon’s most annoying arguments in front of all half million of his Twitter followers. Which really pissed me off.

The Quincy Institute, one of Washington DC’s very few worthwhile think tanks was talking about President Eisenhower’s point that the military industrial complex robs the American people of things we actually need. And then Yglesias decided to weigh in with “The military was over 8 percent of GDP when Eisenhower made that warning!”

I really hate this argument. Yglesias’s unspoken point is that US military spending is supposedly only 3.3% of GDP now, down from 8% in the 1950s so the Quincy institute shouldn’t be complaining. The reason I hate this argument, is that on one level it’s true.

Here’s a graphic from the Defense Department making this case back in 2019. It’s likely that Much more of the US gross domestic product went to defense spending during the heights of the Cold War, when we were fighting very hot wars in Korea and Vietnam. This 3% number is way too low BTW. It doesn’t include what we pay on Veterans healthcare, and it doesn’t include the now almost half a trillion dollars we spend every year on interest payments on the national debt, 6 trillion of which was generated by the Pentagons pointless, lost wars in Iraq and Afghanistan. The Pentagon famously can’t pass an audit, so even the generals don’t know what they are actually spending. But the trajectory represented in this graph is probably accurate . Even if the real number of what we spend on war in 2023 is 4 or 6% of GDP it is still probably lower than it was under Eisenhower in the 1950s.

But that’s nowhere near the whole story. This is one of those statistical data points that’s True-ish, but also fundamentally misleading. Because it leaves a whole hell of a lot out. Like, Oh, I dunno, the entire US social safety net.

Social Security & Healthcare Boom

The United States is often imagined to not have any sort of welfare state. Capitalism red in tooth and claw, that’s us right? Well no, not so much. Poor and young people truly are kind of fucked in the United States, but we actually spend a lot of money on old people. Even the few poor people who last long enough to make it to the age of 67 are relatively decently taken care of. Social Security is a whole mess of programs, but fundamentally it’s a pension for old folks. It was signed into law in August 1935.

In 2022, Social Security was reckoned to cost 4.8% of GDP. That number isn’t static, it’s growing steadily, and it will continue to do so. Social security cost a lot less when it started back in the 1930s. Back then everybody smoked, carbs were considered the heart of a heathy diet, and most kinds of cancer were an instant death sentence. Most people were expected to get a few years of social security. Now there are millions of people who have been receiving benefits for decades. That’s a great thing by the way. A lot of people talk about this like it’s some kind of government failing. It’s not, it’s amazing, but it does create a budgeting math problem that needs to be sorted out.

Now when you look at this pentagon graph considering that math, things look a bit different. In 2022, Social Security costs the federal government 4.8% of GDP. That’s not what it cost in the 1960s when we could afford to blow 9% of GDP on losing the war in Vietnam. That’s not even what social security cost in the 1980s, when Reagan was blowing 5% of GDP on missile defense programs that still don’t work 40 years later. There are just a lot more old people now than there were in the 1960s and the 1980s. And that’s great, but we haven’t even started talking about healthcare yet. In 1965, LBJ enacted Medicare to cover the health of the elderly and the smaller Medicaid program to help out a slice of the very poor. Like social security these programs started out small, but grew steadily. Richard Nixon’s war on Cancer hasn’t been won, but it’s made incredible progress. I have family members who would not be alive today without it. Those cancer drugs aren’t cheap though. All the diabetes generated by our agriculture subsidies isn’t cheap either. In 2021, Medicare and Medicaid together amounted to 7% of US GDP.

Social Security and the healthcare programs represent a choice as to how the US government spends its money. There are many ways to make these programs cheaper, with Medicare especially, but they are never going get cheap enough for Washington, DC to do another Vietnam war, or even another war in Iraq. These pension and health programs are already significantly more expensive than they were in 2003 when George Bush destroyed the rules based order by invading Mesopotamia. Is it such a bad thing that our current financial picture makes it harder for us to jump in to losing wars? And the truth is, if we really needed a war, we could totally afford it. The United States has immense capacities. We actually totally could spend 12% of GDP on these programs for old folks, and spend another 10% forcing China into a war over Taiwan. All that it would require is…


In this video I have provided a rough picture of a US government that has steadily been spending more and more of our gross domestic product on domestic needs over the past century. Taxes are the other side of this question. If the federal government was taking in a higher percent of GDP with taxes, then these choices we have made would be sustainable. But if you know anything about the Reaganite focus of our government for the past 45 years or so, you know that has not happened. In fact, the percentage of national production captured by taxes has remained static… at best.

The typical way to talk about this is to say back in the 50s we made the rich pay much higher taxes, top marginal rates were like 90%. But any tax or legal professional with a sense of history knows that this data point is nonsense. There’s a ton more to tax receipts than income tax, and there are many, many ways to organize a rich person’s life so as to avoid high tax rates like this.

Back before the 1980s tax reform there were many more ways to shelter income with real estate transactions than there are today. And devices as simple as the company car and the expense account could easily give executives a rich person lifestyle that never showed up as personal income, while also lowering the tax burdens of their companies as well. Almost nobody was paying 91% of their income to the government. 1950s income tax rates are one of those true but useless data points, very similar to the defense spending to GDP ratio Yglesias is pushing. I sometimes suspect that the transparent stupidity of the main pro-tax argument in the mainstream media is intentional. The fact that it’s easy to dismiss by anybody who knows anything about the issue makes most arguments for higher taxes seem silly and uninformed.

A better way to argue for higher taxes is to just use the real numbers. The Federal Reserve Bank of St. Louis provides a record of annual tax receipts going back to the 1940s as well as an estimate of GDP.

This graph jumps around a bit. In some periods, like the 1990s, the returns to investment go up too quickly for tax avoidance strategies to keep up with, and the percentage of taxes captured by GDP briefly rises. This is especially true for 2020 and 2021, when a bunch of Reddit powered day traders made a ton of crypto and meme stock money without understanding how capital gains taxes work. That one time bonus is over, and revenue is already crashing back down to earth. The broader trajectory has been clear. Back in the 50s when we spent 8% of GDP on defense and very little on health & pensions, we captured 18-20% of GDP in taxes. Now it’s more like 13-15%. Tax cuts, deregulation, and perhaps most importantly the defunding of the US’s tax collecting bureaucracy have had a massive impact. Social Security and Medicare use most of the money we actually take in every year. That 3% of GDP we spend on all these wars we lose is all borrowed money. And it has been for decades.

Any serious reform of the United States will require tax hikes, but we aren’t there yet. Until we get there, we are making trade-offs. We have been hearing a lot about the debt limit and the US budget deficit out of Washington DC recently, but it’s been absolutely meaningless. There are a ton of things we could do to make both defense and healthcare spending dramatically more effective and cheaper. Mostly by breaking up the monopolies that currently provide the bulk of these services. But the Republicans who recently held the world financial system hostage over the debt refused to even discuss Social Security, Medicare or taxes, and even required that we spend more on defense expenditure. A few handfuls of billions of dollars have been saved by making the lives of the poor a little more miserable, but none of the budget basics described in this video have been impacted in anyway. They were made a little bit worse actually, because the deal takes even more money away from the IRS, making it even harder to raise revenue.

So let’s put it simply. People like Yglesias who are implying that there is all kinds of spare money lying around for more intensive new cold warring couldn’t be more wrong. Our government has real responsibilities now, I am sorry, but we just don’t have the spare cash for world war III anymore.

Before we close, it’s worth pointing out just how uniquely fiscally irresponsible the Pentagon is. If we give money to healthcare in the right ways, it’s possible to imagine improving outcomes and even saving money. The best possible outcome for Pentagon spending is more piles of dead people. The United States hasn’t fought a worthwhile war since Korea. The results have mostly been losses, and US use of force has mostly made the world dramatically worse. The obscene amount we spend on weapons makes these loser wars more likely. That’s not some malicious lefty theory. It’s documented by the statements and actions of US government officials like Madeline Albright. Perversely the more money we give to the Pentagon, the more dangerous the world gets, requiring more money for the Pentagon. This self-eating shit snake of fiscal Armageddon has now made it into official national security policy. Since 2018, the stated policy of the US government has been to go out and pick fights with the handful of countries that waste as much money on weapons as we do. This “great power competition” policy has already lead to devastation in Ukraine, and our policymakers are constantly trying to do the same in Taiwan and in the Persian Gulf. Why in god’s name would we dedicate more of our GDP to this pursuit of war? We should be looking for ways to spend less than 3%, not more.

Now Yglesias claims that he’s just describing reality with his sophomoric commentary. He brings up fricking David Hume to claim he’s just laying out the facts man. But that’s not what he does with his tweet. The Quincy Institute points out that every defense dollar spent is sacrificing some other priority. Yglesias attempts to counter that by saying we’re wasting less of our national treasure on the Pentagon than we did back in the 1950s, so the Quincy institute is somehow exaggerating the trade-off. But that’s exactly wrong. Quoting just the GDP percentage that goes to defense obscures the fact that there’s infinitely more at stake now than there was in the 1950s or the 1980s. Every extra dollar sacrificed to making bombs isn’t taking money away from the school of some poor kid you don’t care about. It’s taking years off your mother’s life. It’s a cancer treatment none of us will get, so a handful of generals can get 7 figure jobs at Boeing. It’s your town’s hospital closing so we can back up some doomed Persian Gulf monarch for another six months. Those are the kinds of trade-offs we’re talking about here. That’s what you should be thinking about whenever somebody tries to tell you we spent more on defense in the 1950s.