Who is the national debt actually owed to, and how is it unlike a household’s?

The debt your grandchildren supposedly owe

Few numbers are quoted with as much dread as the national debt. It has its own ticking clock, it is measured in tens of trillions, and it arrives in political speech attached to a particular fear: that it is a bill run up by the present and left for children not yet born, who will spend their working lives paying off what their grandparents spent. The figure is real and it is large. What the fear gets wrong is almost everything about what kind of thing the figure is, starting with the simplest question, the one the dread skips over. If the country owes all this money, who exactly is the country in debt to.

Where your taxes actually go, and why the money seems to buy so little of what you can see, is a question this volume shares with the chapter on why subsidies outlive their industries (Frameworks, on concentrated benefits and diffuse costs). This chapter takes the debt itself: who holds it, why a government’s borrowing is a different animal from a family’s, and what the historical record says actually happens to debts this size. The short answer to the first question rearranges the fear considerably. A great deal of the debt is the government owing money to itself and to its own citizens.

Owed to whom

Take the United States, whose debt is the one most often invoked as a warning. At the end of 2023 the gross federal debt stood at about $34 trillion, and the Treasury keeps a running estimate of who holds it. The single largest block is not a foreign power or a Wall Street bank. It is the United States government itself. Roughly a fifth of the total is owed to the government’s own accounts, above all the Social Security trust funds, which for decades ran surpluses and were required by law to hold them in Treasury bonds. Another seventh is held by the Federal Reserve, the country’s own central bank, which bought government bonds in the course of doing monetary policy. Together, the government and its central bank hold about a third of the debt: money the state, in effect, owes to itself.

Most of the rest is owed to Americans and the institutions that manage their money. Mutual funds, banks, insurance companies, and pension funds hold Treasury bonds because they are the safest asset a saver can own, the thing a retirement account reaches for when it wants a sure return. State and local governments hold them; individuals hold them directly. Add these domestic holders to the government’s own accounts, and roughly three quarters of the debt is owed at home. The foreign share, the part that belongs to other countries’ governments, central banks, and investors, is a little under a quarter, and its largest single lenders are Japan, China, and the United Kingdom, in that order.

The picture is not unique to the United States. Every large economy that borrows in its own currency shows the same broad shape: a debt owed overwhelmingly to its own institutions and citizens, with the state’s own accounts and central bank holding a large slice. Japan, whose debt is more than twice its yearly output, owes almost all of it at home. The creditor that the fear imagines, a foreign power holding the nation hostage, is a minority holder of a debt the nation mostly owes to itself. That does not make the debt costless. It changes what the debt is.

The household that a country is not

The credit-card image behind the grandchildren fear assumes that a government borrows the way a family does. A family is a currency user: it earns money it cannot create, its debts come due on fixed dates, and it must clear them before the earners die, because a household has an end. On every one of those points a government that issues its own currency is a different kind of borrower, and the differences are what the household analogy quietly drops.

The first difference is issuance. A currency-issuing state does not have to find the money it owes in its own currency; it is the source of that currency (Chapter 2, on where money comes from). This is why such a state cannot be forced into the kind of default a household faces, where the money simply is not there. It can always create what it owes. The catch is not solvency but what that creation does to prices, a point returned to below. The distinction is sharp: a family, a company, a city, or a country that borrows in a currency it does not control, as Greece did in euros, is a currency user and can run out of money; the United States, the United Kingdom, and Japan, borrowing in their own currencies, cannot in the same way.

The second difference is repayment. A household loan amortises: it is built to be paid down to zero. Government debt is not. A maturing bond is almost always paid off by selling a new one, a practice called rolling the debt over, and the total is never meant to reach zero any more than a large company’s borrowing is. A state has no retirement date at which the books must be cleared, because it does not die. The relevant question is therefore not whether the debt is ever repaid, which it is not, but whether it can keep being rolled over on manageable terms, which depends on interest rates and on the size of the economy standing behind it.

The third difference is the one that does most of the work, and it turns on the gap between the interest rate on the debt and the growth rate of the economy. If an economy grows faster than the interest it pays on its borrowing, a condition economists write as the interest rate being below the growth rate, then the debt shrinks relative to the economy even if not a penny of it is repaid, because the denominator outruns the numerator. In his 2019 address to the American Economic Association, the economist Olivier Blanchard pointed out that for most of the past this condition has held: safe government interest rates have usually run below growth rates, which means debt can often be carried, and outgrown, at little fiscal cost. He was careful not to say the debt is free, and the argument reverses if interest rates climb above growth. But it dismantles the intuition that a debt must be actively paid off to stop being a burden. Growth can do the work that repayment never does.

History’s great debts were outgrown, not repaid

That claim is not a modern hope; it is the plain record of the last three centuries, and Britain offers the longest clean run of it. The United Kingdom has twice carried debts larger than the ones that frighten people today, and both times the debt came down not because it was paid off but because the country grew and inflated its way out from under it.

Read the two peaks against what happened to the debt in pounds. After the wars against Napoleon, British debt reached about 194% of national income in 1822. Over the following eighty years the debt in pounds barely fell at all, yet by 1900 it had dropped to about 35% of income, because the industrial economy underneath it had grown perhaps fivefold. The debt was not repaid; it was outgrown. After the Second World War the debt was larger still, about 259% of income in 1946. This time the debt in pounds did not hold steady; it rose nearly sixfold over the next forty-four years. Yet by 1990 it had fallen to about 23% of income, because postwar growth and a long stretch of inflation shrank the real weight of the old borrowing even as new borrowing piled on top. The United States came out of the same war on the same path, its wartime debt eroded by the growth and mild inflation of the decades that followed rather than by any campaign of repayment. The great debts of the past were not paid back. They were grown past, and in part inflated away.

The two warnings worth taking seriously

None of this makes government debt harmless, and the subject itself sits between two confident positions that both contain something true. Each deserves to be stated at its strongest before it is weighed.

The first position pushes toward calm. A school of thought known as Modern Monetary Theory, argued most prominently by the economist Stephanie Kelton in The Deficit Myth (2020), holds that a government issuing its own free-floating currency can never be forced into involuntary default, because it can always create the money to pay a debt denominated in that currency. On this view the entire framing of national solvency is a category error imported from the household: the question is never whether such a state can afford its debts in money terms, since it always can, but whether spending outruns the real resources of the economy, at which point the constraint shows up as inflation rather than bankruptcy. The strong form of the argument is that the deficit is the wrong thing to fear; inflation and real capacity are the right things.

The second position pushes toward caution, and its most influential statement is also its most cautionary tale. In 2010 the economists Carmen Reinhart and Kenneth Rogoff published a study, “Growth in a Time of Debt,” reporting that when a country’s government debt rises above 90% of GDP, average economic growth turns slightly negative, around −0.1%. The finding arrived as governments were debating austerity, and the 90% threshold was cited around the world as a hard line not to cross. Three years later, a graduate student, Thomas Herndon, working with Michael Ash and Robert Pollin, obtained the original spreadsheet and found the result did not hold up. A formula error had omitted several countries from the average; some available data had been excluded; and the countries had been weighted in an unusual way. Corrected, the average growth rate for high-debt countries was not −0.1% but about +2.2%, and the cliff at 90% vanished. What survived the correction was milder and more qualified: high debt and slow growth do tend to travel together, but there is no magic threshold, and the causation runs both ways, since a weak economy raises the debt ratio just as a high debt ratio may weigh on growth.

Hold the two positions together and the real shape of the risk appears. A currency-issuing state does not face a household’s bankruptcy, as Modern Monetary Theory insists; but it does face two dangers the calm version can understate. One is inflation: paying for spending by creating money, past the point where the economy can supply the goods, devalues the currency and every claim denominated in it, which is the collapse case in Volume V, on what happens when money dies. The other is refinancing: a government must keep rolling its debt over, and if lenders lose confidence and demand punishing interest rates, or if the debt is owed in a currency the state cannot issue, the rollover can fail. That failure, when concentration of wealth meets a state that cannot tax or refinance, is the fiscal collapse that ends a regime, told through the French crown in Volume V, on when inequality breaks a regime. The debt is not a bill your grandchildren must clear. It is a standing arrangement whose real risks are inflation and lost confidence, not insolvency, and those risks are governed by the same monetary machinery the previous chapter described.

Where the argument goes next

Seen plainly, the national debt is less a loan the country took from someone else than a web of claims the country holds against itself, carried forward indefinitely and shrunk over time by the growth of the economy underneath it. The fears that attach to it are not baseless, but they point at the wrong hazards: the real ones are inflation and lost confidence in the rollover, not a due date at which the grandchildren are handed a bill. Reading the debt this way depends on everything the last two chapters set out, that money is bank-created credit (Chapter 2) and that a central bank stands behind it and steers its value (Chapter 3).

This volume has now followed money from its origin as a ledger entry, through its creation by banks, to the public institutions built to manage it. The institutions that come next are private ones, and the first is the most consequential invention in the history of business: the company that can outlive its owners, pool the savings of strangers, and shield them from the debts it runs up. How the corporation was assembled, and why it behaves like a piece of technology, is the next chapter’s subject, when this book takes it up.