Debt: A Primer

What Debt Is

A person who lends money is a creditor. Creditors give up present possibilities and postpone their own spending into the future. A person who borrows is a debtor and does the reverse: he acquires possibilities now and commits future income to paying for them. In nearly all cases the debtor repays the principal plus interest.

The interest rate is negotiated between the two parties, and it compensates the creditor for three separate things. The first is the price of exchanging future possibilities for present ones, which would exist even in a world of perfect certainty and stable prices. The second is expected inflation, since the creditor will be repaid in money that buys less. The third is the risk that the debt will not be repaid at all. Keeping these three apart is the beginning of clear thinking about debt. When a government’s borrowing costs rise, the interesting question is always which of the three has moved, because the remedies are entirely different.

A second distinction runs through everything that follows. Debt incurred to finance consumption reduces net worth by definition: the goods are used up and the obligation remains. Debt incurred to finance investment need not, provided the asset acquired is worth at least what was paid for it. This holds for households and for governments alike, and it is the reason a debt figure by itself tells you almost nothing. You have to know what was bought.

Public debt is not simply large private debt

For a private borrower the relationship is plain. There is an identifiable debtor, an identifiable creditor, and a contract. If the debtor cannot pay, the creditor loses money. Public debt looks similar and is usually discussed as though it were the same thing, but the resemblance weakens as the borrowing entity grows.

Consider first a small public body: a municipality, a school district, a public corporation, especially one without the power to tax. These behave very much like private borrowers, because nobody stands behind them. Switzerland provides an unusually clean demonstration. In 1998 the Valais spa town of Leukerbad, with roughly 1,600 residents, was found to owe 346.5 million francs, much of it sunk into thermal baths and a town hall of remarkable ambition. It became the first Swiss municipality ever placed under compulsory administration. Its creditors sued the canton for failed supervision. The Federal Supreme Court accepted that the canton had breached its official duties and held nonetheless that it bore no liability toward the creditors of its municipalities: lenders are free to inform themselves about a borrower’s condition, and doing so is easier with a public body than with a private one. The creditors wrote off 78 percent of their claims. That is private insolvency in every respect that matters.

Now consider the other end: a national government that issues its own currency. Here the analogy quietly fails. Such a government pledges no assets against default, and there is no entity within its own society to which it stands as debtor in the ordinary sense. What we call the national debt is better described as a pattern of claims and obligations among citizens, brokered by the state. The bondholder holds the claim; the future taxpayer carries the obligation; the state sits between them as an intermediary, much as a bank sits between depositors and borrowers without thereby becoming an indebted entity itself. Some economists regard the entire vocabulary of national indebtedness as a category error on these grounds, arguing that qualities are being ascribed to an entity that cannot logically possess them.

You need not accept the strong form of that argument to draw the practical conclusion from it. The useful dividing line is monetary sovereignty. A government that controls the currency in which it has borrowed can always meet its obligations in nominal terms. For such a government the question is never whether it pays, but who bears the cost and through which channel. A government that has surrendered that control is a debtor in the strict sense and can become insolvent. Members of a currency union belong on that second side of the line, along with cantons, American states, and towns like Leukerbad. France and Italy are large countries that borrow in a currency they cannot issue, which places them closer to Leukerbad, in this one respect, than to Japan.

Who carries the burden

If a private person borrows, he normally repays later out of his own income. Should he die first, his heirs inherit the obligation, and they are free to refuse the inheritance. Nothing of the kind applies to a political community. When the state borrows, a diffuse transfer takes place from future taxpayers to present ones, and its size depends on how long the debt runs and how far the borrowing financed genuine investment rather than consumption.

Two points here are commonly misunderstood. It does not matter for the burden on future taxpayers whether the state borrowed at home or abroad; the taxpayer of 2045 must raise the same sum either way. What differs is where the money then goes, since interest on external debt leaves the country while interest on internal debt is redistributed within it. And the popular idea that internally held debt burdens nobody because we owe it to ourselves rests on adding up the national balance sheet and observing that claims and obligations cancel. They do cancel in the aggregate. They do not cancel for the individuals concerned, and it is individuals who pay taxes.

How to Judge the Size of a Debt

Four criteria are in general use. Present value: what is the discounted value of the benefits still expected from whatever the borrowing paid for? Opportunity cost: what price would the underlying asset fetch if sold today? Net worth: how large is the debt relative to total assets, and is net worth positive after the debt is deducted? Serviceability: what share of current income must go to interest and amortization, bearing in mind that the interest rate can move?

A private borrower can be assessed on all four, and in practice lenders rely on the last two. Public debt is different, because only serviceability survives contact with reality. Individual benefit streams from public spending are poorly known. Even where they are known, they cannot be meaningfully aggregated across persons. And governments hold a great many long-lived capital assets whose value is genuinely obscure, from road networks to military hardware to school buildings. None of this stops politicians from justifying new borrowing by its supposed benefit to the community. It does mean that such claims cannot be checked, which is a reason to discount them rather than to repeat them.

Serviceability itself comes in four versions.

Serviceability I asks what share of current tax revenue must go to interest. Serviceability II compares net debt to GDP; the European convergence criterion sets 60 percent of gross debt. Serviceability III compares annual new borrowing to GDP, with a European criterion of 3 percent. Serviceability IV asks how net debt will develop relative to projected GDP.

The first of these is the most informative, and three things deserve attention when using it.

The rate at which public bodies borrow fluctuates, and small absolute movements produce large relative ones. A move from 0.5 to 1.5 percent triples the interest bill even though it is only one percentage point. Second, the effect of any such move scales with the outstanding stock: the same one-point rise is trivial at a low debt level and severe at a high one. Third, once the interest rate exceeds the rate of nominal economic growth, the interest bill grows faster than the base from which it must be paid, assuming a constant tax ratio. That is the point at which arithmetic turns against the borrower and a primary surplus becomes necessary simply to stand still.

Projections of the fourth kind rest on three drivers, and separating them is more useful than any single headline number. The first is the primary balance, meaning revenue minus spending before interest. The second is the relationship between the interest rate and nominal growth. The third is one-off stock adjustments such as bank rescues or privatizations. Politics decides the first. The other two happen to it.

Three traps in international comparison

Debt figures are compared across countries far more often than they are defined, and three distinctions matter enough to check every time.

The first is the level of government. American federal debt stood near 122 percent of GDP in gross terms in early 2026, but state and local borrowing adds roughly another 13 percentage points, bringing general government debt closer to 129 percent. In Switzerland the distortion runs the other way and is larger in proportion: federal net debt was 16.1 percent of GDP at the end of 2025, while cantons and communes together owe about as much again, putting the general government figure near 25 percent. Anyone comparing the Swiss federal number with the French general government number is comparing two different things. The same trap catches serviceability I: American interest costs consume about 19 percent of federal revenue, but federal revenue is only 17.5 percent of GDP, whereas European figures are measured against general government revenue of 40 to 50 percent of GDP. The American ratio looks worse partly because its denominator is narrower.

The second is gross versus net. Japan’s gross debt exceeds 200 percent of GDP; its net debt is roughly half that, because the state holds enormous financial assets and the central bank holds a large share of the bonds. Denmark and Norway are similarly flattered by netting. Net figures are the more meaningful measure and the less consistently available one, which is why gross ratios continue to dominate public discussion.

The third is the definition of debt service. Japan’s fiscal 2026 budget sets aside 31.3 trillion yen for what is translated as debt servicing, but that figure includes principal redemption under a statutory sixty-year rule. It is not comparable to an interest bill.

One further caveat applies to all ratios measured against GDP. Where the state absorbs half of national output, GDP is a poor yardstick, because a large part of the denominator is itself the product of public spending.

Where Today’s Public Debt Is Heading

Five paths are open to a heavily indebted state, and they are best understood as five answers to a single question: who ends up paying?

Continuation without a break. The ratio rises, interest crowds out other spending, and nothing snaps. No decision is required; this is the sum of decisions not taken, which is why it is usually the single most likely path.

Erosion through inflation and financial repression. The real value of the debt is reduced at the expense of bondholders. This was the method by which the Western world worked off its war debts between 1945 and 1955, and it requires both monetary sovereignty and a captive investor base.

A market event. A failed auction, a jump in the term premium, a loss of confidence, followed by correction under duress. The correction is then badly composed, because it is made in a hurry.

Orderly consolidation. A sustained primary surplus achieved through spending restraint and higher taxes, decided in advance rather than imposed by markets.

A growth surprise. Nominal growth durably exceeds the interest rate and the ratio falls without anyone being asked to sacrifice anything. This is the only painless exit and historically the rarest. In the great nineteenth-century consolidations in Britain, France, and the United States, the interest rate exceeded the growth rate throughout, and the entire work of debt reduction was done by primary surpluses.

How the major borrowers stand

The United States crossed 40 trillion dollars of gross federal debt in August 2026, with about 32 trillion held by the public, roughly 100 percent of GDP. Interest costs passed one trillion dollars in 2026 and now absorb about 19 percent of federal revenue, up from 9 percent in 2021 and projected to reach a quarter by 2036. The average rate paid on the existing stock is still below nominal growth; the rate on new issues is not, so the average is climbing as old low-coupon debt matures. Consolidation is politically improbable, because the federal tax base is narrow: there is no national sales tax, and a large share of households owe no federal income tax, so the median voter does not feel the cost of federal spending directly.

The United Kingdom carries public sector net debt near 95 percent of GDP and spends about 3.6 percent of GDP on debt interest, among the highest levels in fifty years. Two structural features make it unusually sensitive. Around a quarter of the stock is index-linked, so inflation raises debt service immediately rather than eroding the debt, closing the usual escape route. And interest is paid on central bank reserves at the policy rate, which is floating-rate debt with no maturity. Britain also demonstrated in September 2022 how quickly confidence can go.

France carries about 118 percent of GDP with a deficit above 5 percent and a primary deficit that no current plan closes. It has been downgraded three times in twelve months and its debt path points toward 130 percent by 2030. It cannot inflate, and its political system has so far proved unable to pass a serious consolidation.

Italy carries the highest ratio of the group at roughly 138 percent and has the best current trajectory: a primary surplus, a deficit back below 3 percent, and a risk premium at a fifteen-year low. It is the clearest available demonstration that the level of debt tells you less than the primary balance does.

Japan has gross debt above 200 percent, most of it held domestically and a large share held by its own central bank. Its ten-year yield touched 3 percent in September 2026 for the first time since 1996, and the finance ministry has raised the interest rate assumption underlying its budget to a twenty-nine-year high. Japan has the greatest institutional capacity for financial repression of any large economy and the least political appetite for consolidation.

Two well-governed small states serve as controls. Switzerland shows general government debt near 25 percent, a structural surplus in 2025, and the first overall financing surplus since 2019. Denmark shows about 29 percent and ran a general government surplus of 2.9 percent of GDP in 2025, one of only five European countries in surplus. Neither is exempt from demographic pressure, but neither faces a debt problem in any meaningful sense.

The distribution of likely outcomes follows monetary sovereignty rather than the size of the debt. Inflation is available to Japan, the United States, and, in a limited way, Britain. It is not available to France or Italy, for whom the pressure must therefore escape through prices in the bond market or through politics. That is why the market-event path weighs more heavily for the euro members than their headline numbers alone would suggest, and why an outright restructuring, which is a remote possibility for a currency-issuing state, is a real branch for a member of a currency union. Greece in 2012 is the precedent.

What Can Be Done

Fiscal rules work, and they leak. Switzerland demonstrates both halves of that sentence at once. Its constitutional debt brake, in force since 2003, is credited by the federal finance administration with the country’s low debt level, and an independent estimate against a control group of comparable countries suggests that without it the Swiss debt ratio would have risen by two to three percentage points every year, leaving a debt stock roughly three times its actual size. At the same time, the deadline for repaying pandemic-era debt has been extended to 2035, the most recent relief package failed to close the structural gap for 2027, and additional measures had to be decided in the spring of 2026. The rule is not broken; it is stretched. The same pattern appears in the United States, where forty-nine of the fifty states operate under balanced budget requirements that apply to operating budgets and are routinely circumvented through capital borrowing and unfunded pension promises. A binding budget constraint also tends to push spending into regulation, which achieves similar ends without appearing in any budget.

Visibility matters more than the rule itself. A fiscal rule’s real function is to prevent the cost of public spending from being deferred into debt, keeping it in front of the current taxpayer. Whether that visibility produces political pressure depends on how broadly the tax burden is spread. Where taxation reaches the middle of the income distribution through broad-based instruments, voters experience the cost of the state directly, and the demand for consolidation has somewhere to come from. Sweden after 1994 and Denmark today are examples. But the mechanism is not automatic, and two counterexamples should temper any confidence in it. France has one of the highest tax ratios in the developed world and has been unable to consolidate for two decades. Belgium reduced its debt from 133 percent of GDP in 1993 to under 90 percent by 2007 without a currency of its own, then let it climb back above 107 percent. Consolidation is achievable and it does not stay achieved by itself.

Starving the beast does not work. The theory holds that cutting taxes will force spending discipline. It rests on a naive picture of how governments behave. In practice the present generation can shift the shortfall onto later ones for a very long time, and generally does.

Growth is the best answer and the least reliable one. No government can decide to grow faster, and the historical record shows that debt has almost never been retired by growth alone.

A closing observation, which is really the point of the exercise. The question people usually ask about public debt is how much is too much, and it has no general answer. The ratio that is comfortable for one country is dangerous for another, depending on the currency it borrows in, the breadth of its tax base, the maturity of its obligations, and the credibility of its institutions. The better question is the one the five paths are five answers to: when the bill comes, who pays it? Bondholders through inflation, taxpayers through consolidation, beneficiaries through cuts, or nobody, because growth arrived. That question always has an answer, and it is a question about politics rather than arithmetic.

Discussed and written with Claude.