When the debt
writes monetary policy.

Think of a household paying its mortgage interest every month whether or not it renovates the kitchen: the renovation is discretionary, the interest is not. Governments split their budgets the same way.

Every government lives inside one accounting identity. In a given year it collects revenue, mostly taxes, and it spends. Spending splits into two very different pieces: the goods, wages, and transfers it buys to run the country (call this primary spending), and the interest it owes on debt it already carries. The gap between revenue and primary spending is the primary balance. Subtract interest too and you get the total balance. When either is negative it is a deficit.

Interest is the dangerous term because it is not a policy choice each year: it is the interest rate times the debt you already owe. A larger debt means a larger interest bill, a larger total deficit, more borrowing, and a still larger debt next year. That feedback loop is the whole story of Part II. For now, build a budget and watch the interest bill drive a wedge between the primary and total balances.

A deficit has to be paid for somehow. Once you have decided to spend more than you tax, the identity that follows is unavoidable: the gap is closed either by borrowing (selling bonds to the public) or by money creation (the central bank credits new base money to the treasury). Written as a flow, the deficit equals the change in bonds outstanding plus the change in base money.

These two financing routes look interchangeable on the page, but they are not. Bonds are a promise to repay in real terms; they add to the debt stock and to next year's interest bill, but they do not directly create inflation. Money creation adds nothing to the debt the government must service, but it expands the money base, and past a point that is inflationary. The central question of fiscal dominance is which lever gets pulled, and who decides. Slide the financing mix and watch the two consequences trade off.

Nobody worries about a government owing a trillion in the abstract; they worry about whether it can service that debt out of its economy. So the number that matters is not the debt in dollars but the debt relative to GDP, the ratio d = B / Y. A debt is sustainable when the economy it draws on is large enough, and it grows or shrinks not just when B changes but when Y changes underneath it.

This is the quiet reason the denominator matters so much. Even with the nominal debt frozen, a growing nominal GDP (from real growth plus inflation) shrinks the ratio every year. A country can "grow out of" debt without repaying a cent, and inflation helps because it inflates Y while B, fixed in nominal terms, stays put. Freeze the debt and watch nominal growth alone melt the ratio.

Suppose a country owes exactly the size of its own economy: debt equals 100% of GDP. It pays 5% real interest and grows 2% a year, with no primary balance at all. Next year it owes 105 against an economy that has grown to only 102, so even without borrowing another cent, the ratio ticks up. That single example is the whole law of motion; here is the general version.

Now put the pieces together into the single equation that governs everything that follows. From one year to the next, the debt ratio grows by the real interest rate (interest compounds the stock), shrinks by real growth (the denominator expands), and falls by whatever primary surplus the government runs to pay debt down. Writing d for the debt ratio, r for the real interest rate, g for real growth, and pb for the primary surplus, the exact recursion is below.

This one-line difference equation is the heart of the guide. Everything about sustainability is encoded in the multiplier (1+r)/(1+g). If it exceeds 1 (that is, if r > g), the debt ratio is pushed upward every year by its own weight, and the process is explosive unless a primary surplus offsets it. If it is below 1 (r < g), the debt ratio is self-correcting and converges to a finite steady state, even with a persistent primary deficit. Set the four dials, then step the economy forward year by year and watch which regime you are in. The amber line is the steady state the honest algebra predicts.

Picture an actual snowball rolling downhill. It does not add a fixed scoop of snow on every turn; it picks up more precisely because it is already bigger, so it grows faster the farther it rolls. A debt under r > g behaves the same way: this year's interest is a percentage of a balance that already includes every past year's interest, so it compounds faster the larger it gets.

Strip out the primary balance and look at what the debt does on its own. With pb = 0, the change in the ratio each year is the snowball term: the growth-adjusted interest cost, ((r − g)/(1+g)) · d. When r > g this term is positive and proportional to the debt itself, so a bigger debt snowballs faster, the signature of an unstable feedback. When r < g it is negative: growth melts the ratio down faster than interest piles it up, and the debt shrinks even while the government runs deficits.

Set the rate and growth and read the snowball two ways: the one-year push on the ratio, and the pure snowball path over a decade with no primary balance at all. The amber marker sits at the knife edge, r = g, where the snowball is exactly still.

Think of running up a downward-moving escalator. You need a certain pace just to stay in the same spot, and that pace depends on how fast the escalator moves and how far back you have already slid. A country with r > g is on that escalator: its debt ratio keeps sliding unless the primary balance supplies exactly enough pace to hold position.

If the snowball pushes the ratio up by ((r−g)/(1+g)) · d each year, then to hold the ratio flat the government must run a primary surplus of exactly that size. Setting the change to zero gives the debt-stabilizing primary balance, the surplus that neither pays debt down nor lets it grow.

Read this equation and the politics of fiscal dominance fall out of it. When r > g, a heavily indebted country needs a persistent, large surplus just to stand still, and the required surplus rises with the debt itself. If that surplus is politically impossible, the debt drifts up until something gives. When r < g, pb* is negative: the country can run a permanent primary deficit and still keep its debt ratio stable. Compare the surplus you actually run against the surplus the arithmetic demands.

The printing press is not just an escape hatch; it is a source of revenue with a name. Seigniorage is the real resources a government captures by issuing new money. Each period it prints money growing at rate μ, and because people hold real money balances m = M/P, the new money buys real goods worth roughly money growth times real balances.

Viewed from the other side, the same flow is the inflation tax. New money pushes the price level up at rate π, which quietly erodes the real value of every unit of money the public already holds. The base of this tax is real balances m; the rate is inflation π. In a steady state where real balances are constant, the two views coincide: seigniorage collected equals the inflation tax paid, μ · m = π · m. For now, hold real balances fixed and see that revenue looks temptingly linear in how fast you print.

The straight line of the last chapter was a trap, because the tax base is not fixed. The faster you print, the higher inflation runs, and the less real money anyone wants to hold: people spend cash faster, switch to foreign currency, or barter. Money demand falls with expected inflation. A standard and empirically grounded form is Cagan's: real balances m = m₀ · e−απ, where α measures how sharply holdings shrink as inflation rises.

Put that shrinking base into the seigniorage formula and revenue becomes R(μ) = μ · m₀ · e−αμ. It rises, peaks, then falls: the seigniorage Laffer curve. Print too fast and the collapsing base outweighs the higher rate, so real revenue actually drops. The revenue-maximizing money growth is exactly μ* = 1/α. Past that peak lies the hyperinflation trap: governments print ever faster to chase revenue that is falling. Sweep the rate and watch revenue climb and then break.

Picture two drivers heading toward each other on a single-lane road. The treasury sets how large a primary surplus it will deliver; the central bank sets how much inflation it will tolerate. Between them they must cover the same bill, so if one refuses to move, the other has to swerve.

The government budget constraint and the debt law of motion must both hold at once, which means the primary balance, the interest-growth snowball, and seigniorage are chained together. To stabilize the debt, the snowball ((r−g)/(1+g)) d must be covered by the primary surplus plus seigniorage: pb + s = ((r−g)/(1+g)) d. Two authorities set two of those terms, and something has to give. Which one adjusts defines the regime.

Flip the regime and see who is on the hook. Under monetary dominance the required primary surplus is the endogenous answer (in amber); under fiscal dominance, with the surplus pinned down, the required inflation is the answer instead, backed out from the inflation tax π = s / m. Raise the debt or cut the surplus and watch the required inflation climb.

In 1981 Thomas Sargent and Neil Wallace showed something that still unsettles people: if the fiscal authority will not budge, a central bank that fights inflation with tight money today can cause higher inflation tomorrow. The logic is pure accounting. Tight money means less seigniorage now, so the unfunded deficit is covered by issuing more bonds. That extra debt carries interest. When borrowing finally hits its limit, the central bank must monetize a debt that is now larger, and it takes more inflation to do it.

The widget runs the exact two-phase experiment. In the tightening phase the central bank holds inflation low, so seigniorage is small and the debt snowballs under the deficit. When the phase ends, the debt must be stabilized at its now-higher level, and the required inflation is backed out from the inflation tax. Tighten harder, or tighten longer, and the eventual inflation (in amber) climbs. That is the trap: the honest arithmetic refuses to let the deficit disappear.

There is a more radical way to see the same forces, called the fiscal theory of the price level. Take the government's nominal debt B as given, and ask what it is really worth. The theory says the price level P adjusts so that the real value of the debt equals the present value of all future primary surpluses that will ever be raised to back it. Debt is like equity in the government: its real value is the discounted stream of what stands behind it.

The unsettling implication: if markets come to expect smaller surpluses, the price level jumps up on its own to shrink the real debt back into line, without the central bank doing anything and sometimes against its wishes. Inflation here is not too much money chasing goods; it is a repricing of government debt. Move the nominal debt and the expected surpluses and watch the price level, which is really the market's valuation of the sovereign, respond.

Germany after the First World War is the textbook case, and it fits the arithmetic exactly. Saddled with reparations it could not pay and a parliament that could not agree to tax, the government ran vast deficits and covered them the only way left: the Reichsbank printed. Early on this raised real revenue. But as the last chapters warned, printing pushed the government past the peak of the seigniorage Laffer curve. Germans dumped marks as fast as they earned them, real balances collapsed, and the government printed ever faster to chase a revenue that was shrinking.

The result was compounding gone vertical. By the autumn of 1923 prices were doubling every few days; a wage paid in the morning bought a fraction of its groceries by evening. The famous images of banknotes as wallpaper and wheelbarrows of cash for bread are the visible face of a base that had almost vanished. What finally stopped it was not more printing but a new currency and a credible commitment to stop, the subject of Part VI. Set a monthly inflation rate and watch a fixed pile of cash evaporate as the compounding runs.

The Latin American debt crises of the 1980s show fiscal dominance arriving through the interest rate rather than the printing press. Through the 1970s, countries like Argentina, Brazil, and Mexico borrowed heavily in dollars while growth was strong and real rates were low: the comfortable r < g world where debt melts on its own. Then the environment flipped. The Volcker disinflation sent world interest rates sharply higher, commodity prices fell, and lenders stopped rolling over the debt. A sudden stop.

Overnight r jumped above g and the snowball reversed direction. Debt that had looked stable began compounding uphill, and governments unable to borrow or tax turned to the central bank, delivering exactly the high inflation and repeated devaluations of that "lost decade." The lesson is that the r < g comfort can vanish without warning; sustainability assessed at today's rates can be an illusion. Run a debt path that is calm under low rates, then apply a rate shock and watch the same debt turn explosive.

Not every high-debt story ends in inflation. After 1945 the United States and the United Kingdom carried debt above 100% and, in Britain's case, above 200% of GDP, then brought it down over decades without default and without hyperinflation. They did it with the benign side of the same law of motion: strong postwar growth, real interest rates held low (partly by financial repression), moderate inflation lifting nominal GDP, and steady primary surpluses. Every one of those is a term in the debt equation working in the right direction.

It helps to see any year's change in the ratio decomposed into its drivers. Interest pushes the ratio up; growth pulls it down; the primary balance pushes it down when it is a surplus. When growth beats interest and the budget runs a surplus, all three arrows point down and the debt falls briskly. This is also why today's high-debt advanced economies are not automatically in crisis: with r near or below g, the snowball is a tailwind, not a headwind. Set the drivers and read the yearly change as a stack of contributions.

Here is the deepest reason central-bank independence and credibility are worth so much. Recall the seigniorage Laffer curve: for any revenue below its peak, there are two money-growth rates that raise it, one on each side of the hump. A low-inflation equilibrium where people hold lots of real money and the government prints slowly, and a high-inflation equilibrium where people hold almost none and the government prints furiously. Both raise the same real revenue. Which one you land in is set by expectations.

This is why the escape from fiscal dominance is not only about arithmetic but about beliefs. A credible, independent central bank, a currency reform, a binding fiscal rule: each works by convincing people that surpluses will come and printing will stop, which anchors expectations on μlow. Solve both roots for a required revenue and see the two worlds that share it. Amber marks the good equilibrium credibility buys you; push the revenue need up toward the peak and watch the two collapse together, past which no stable inflation can finance the gap at all.

Trace the whole chain once more. A government's budget must close, so any deficit is financed by bonds or money. Bonds pile into a debt stock whose ratio to GDP obeys one law of motion, driven by the (r−g) snowball and the primary balance. When r > g and the required stabilizing surplus is politically out of reach, the pressure has to escape somewhere. Money creation is the release valve, but it is a tax with a Laffer ceiling, so beyond a point more printing yields less real revenue and only more inflation.

That is the moment of fiscal dominance: the fiscal position, not the central bank's preferences, sets inflation. Sargent and Wallace show the central bank can only choose its timing; the fiscal theory shows the price level revaluing the debt directly; the historical episodes show both the catastrophe (Weimar, the 1980s sudden stops) and the escape (postwar consolidations). And the exit runs through credibility, because the same arithmetic that permits a low-inflation equilibrium also permits a ruinous one, and only belief decides which.

You can now read a budget as a constraint, iterate the debt law of motion in your head, spot the (r−g) regime, price the inflation tax and its ceiling, and recognize fiscal dominance when a headline describes it without naming it. That is the working literacy of sovereign finance.

You have the mechanics. To go deeper, follow the threads that start here:

• Money and banking: where reserves, the money base, and seigniorage actually live on a central bank's balance sheet (the sibling guide in this series).
• Macroeconomics: how the policy rate, growth, and inflation are jointly determined, and how the Taylor rule fits a central bank that still has control.
• Sovereign debt and default: what happens when a country chooses not to inflate but to restructure, and how markets price that risk.
• Central-bank independence: the institutional design, from independence and inflation targeting to the limits both hit when deficits are large.
• The empirical record: cross-country evidence on (r−g), debt reversals, and the real-world shape of the seigniorage Laffer curve.

Every model in this guide ran the real math in your browser. The best next step is to change the numbers until the mechanisms feel obvious, then read the next debt-and-inflation headline knowing exactly which term is moving.