Public Budgets and Fiscal Choices
What makes a government’s promise to repay its debts seem reliable—or risky? France’s reported 2026 borrowing-cost surge offers a striking puzzle: debt levels alone cannot explain why markets charge countries different rates. This course shows that public budgets are both accounting constraints and political choices, and that borrowing costs reflect not only debt and economic growth, but also whether institutions can agree on a plan and who can support the debt in a crisis. You’ll learn how taxes, spending and deficits fit together; what bond yields and spreads signal; and how interest rates, growth and political trade-offs shape debt over time. Using country comparisons and a practical three-part framework, you’ll gain tools to judge fiscal claims without mistaking one headline for a verdict. We begin with the budget itself: what must add up, and who decides how.
Chapter 1: The Ledger and the Promise—What a Budget Is Really Doing
1.1: The Identity That Won't Negotiate: Revenue, Spending and New Borrowing
France's 2026 budget did not fail for want of arithmetic. It failed because the arithmetic had to be agreed. In the winter of 2025–26, rival blocs and a government without a majority turned the finance bill into a test of consent. The numbers were contested, but the deeper dispute was about who would bear the adjustment. That is the right way to enter the public budget: first as an identity that cannot be negotiated, then as a political settlement that must be.
Public budget identity
At its simplest, the public budget identity says that what a government spends must be financed by what it raises or by borrowing. In its core form:
Read it in three steps: subtract expenditure from revenue; call the result the deficit; and treat that deficit as the net new borrowing required. The identity holds after the fact, once the year's receipts and outlays are known. Revenue includes taxes, social contributions, and non-tax receipts. Expenditure includes primary spending—public goods, transfers, administration—and interest payments on existing debt. If revenue exceeds expenditure, the deficit is negative: a surplus, and net borrowing is negative. What must balance is the financing gap: every euro of spending not covered by revenue must be matched by a euro of net new borrowing. #(Stiglitz, 2000, Economics of the Public Sector)
The most useful distinction is between the primary balance and the total balance. The primary balance strips out interest: primary revenue minus non-interest spending. The total deficit includes interest. Past borrowing becomes a present obligation. A government can run a primary surplus and still borrow if interest payments are large enough. The identity links past and present budgets, not just this year's choices.
New borrowing is not an abstraction. It is a sovereign bond sale, a treasury bill, or another debt instrument. The bond market does not merely record the deficit; it prices the risk that future revenue will not cover future obligations. That is why a yield can rise before a budget is passed. #(Cottarelli and Viñals, 2009, A Strategy for Renormalizing Fiscal Policy)
| Term | What it measures | Time dimension |
|---|---|---|
| Revenue | taxes, social contributions, non-tax receipts | flow, over the fiscal year |
| Expenditure | primary spending plus interest | flow, over the fiscal year |
| Deficit | revenue minus expenditure | flow, over the fiscal year |
| Debt stock | accumulated net borrowing | stock, at a point in time |
The arithmetic is neutral; the components are not. Taxes and spending are politically explosive because they distribute gains and losses. Spending often delivers concentrated benefits to identifiable groups while its tax cost is spread across the whole polity—the common-pool problem. #(Weingast, Shepsle and Johnsen, 1981, The Political Economy of Benefits and Costs) Borrowing appears to defer the cost, which is why it can be the path of least resistance. But deferral is not abolition: it shifts the bill to future budgets and future taxpayers. #(Buchanan and Wagner, 1977, Democracy in Deficit)
Analysis: The identity is neutral only in form. In practice, each term is a site of distributional conflict: a tax change names a payer, a spending cut names a loser, and a borrowing target names a future constraint. The same arithmetic can produce opposite political conclusions.
Political institutions shape how that conflict is resolved. Where many veto players—coalition partners, upper chambers, constitutional courts, or external creditors—can block a change, budgets become harder to adjust and deficits can become entrenched. #(Tsebelis, 2002, Veto Players: How Political Institutions Work) This is why governments fail to pass budgets even when the arithmetic is transparent. The identity tells you the size of the adjustment; it does not tell you who pays. A budget fails when consent fails, not when the addition fails.
A deficit is a flow. It is measured over a period, usually a fiscal year. Debt is a stock. It is measured at a moment. The debt stock is the accumulation of past deficits and surpluses:
A government can reduce its deficit and still see its debt rise if the deficit remains positive. It can also run a surplus and still have a large debt stock. Confusing the two is like confusing the water flowing into a bathtub with the water already in it.
The household analogy is tempting, and misleading. A household faces a hard credit limit, cannot set tax rates, and cannot issue a currency. A sovereign can tax, borrow, and—in some monetary regimes—issue money. Its spending decisions also feed back into the economy: cutting spending can shrink the tax base, and tax rises can dampen demand. #(Wray, 2012, Modern Money Theory) But sovereignty is not omnipotence. Bond yields, rollover risk, inflation, and political consent all constrain the budget. The constraint has a different shape, not no shape.
The identity forces a trade-off among three levers: taxes, spending, and borrowing. A government can raise taxes, cut spending, or borrow more. It cannot choose all three independently and expect the identity to hold. If spending rises and taxes do not, borrowing must rise. If borrowing is capped, taxes must rise or spending must fall. The political conflict is over which lever moves and how fast. This is where the identity enters the tax-and-spending analysis in 1.2, and where its accumulated debt stock becomes the debt equation in 2.2.
Conclusion
The identity is the skeleton of fiscal policy. A deficit is a flow, a debt stock is an accumulation, and new borrowing is the bridge between them. The arithmetic cannot be negotiated, even when the politics can. That is the double nature of the budget: a ledger and a settlement. Once you see both, you can ask the right question of any fiscal plan—not simply whether it adds up, but who bears the adjustment, and when.
1.2: The Political Ceiling: Taxes, Spending Claims and the Price of Past Choices
It is the night before the budget is tabled, and the arithmetic is already finished. What remains is the argument. One minister wants the revenue line nudged upward, another insists the spending line cannot move, and a third mutters that the interest bill will be paid whoever wins the room. That is the public budget in miniature: a set of numbers that must reconcile, and a set of promises that must be ranked. Subchapter 1.1 showed that the budget identity only closes if one line gives way; if revenue is politically capped and borrowing is market-priced, expenditure becomes the residual. This subchapter explains why each revenue and spending line carries a different political price, and why a budget can fail even when the sums are clear.
Direct and indirect taxation
A government raises money in two broad ways, and the difference matters more politically than it does arithmetically.
- Direct taxes fall on the income, wealth or profit of a person or firm: personal income tax, corporate profits tax, property tax #(Musgrave & Musgrave, 1989, Public Finance in Theory and Practice). The statutory taxpayer is usually also the person who bears the burden, and the payment is visible as a payslip deduction or an annual return.
- Indirect taxes fall on transactions: value-added tax, sales tax, excise duties, customs duties. The business remits the money, but the burden is designed to travel to the final consumer inside the price #(Stiglitz & Rosengard, 2015, Economics of the Public Sector).
That visibility gap is the political hinge. Because income tax arrives as a deduction, citizens can price it precisely and argue about it loudly. Because a value-added tax is folded into a shelf price, its burden is easier to underestimate — a mild form of fiscal illusion #(Buchanan, 1967, Public Finance in Democratic Process). It is no accident that many advanced economies shifted weight toward consumption taxes in recent decades while holding top statutory income tax rates below their mid-century peaks #(OECD, 2023, Revenue Statistics 2023).
Analysis: Visibility and burden are different things. An indirect tax is quiet at the till and loud in the household budget: it compresses purchasing power at the moment of purchase, and because lower-income households spend a larger share of what they earn, consumption taxes tend to bite hardest there #(Mirrlees et al., 2011, Tax by Design). Calling them painless confuses administrative convenience with economic cost.
Receipts, once collected, do not arrive as a free pool. They land in a queue with a fixed order, and the queue is the part of budgeting that voters rarely see. Three ranks of claims sit on the revenue line.
- Debt service — interest owed on the debt stock, in the identity. It is contractual, due on schedule, and paid before programs.
- Mandatory spending — statutory transfers such as pensions and health entitlements, written into law before the fiscal year begins.
- Discretionary spending — annually appropriated items such as infrastructure, defence and education, which is where the residual actually lives.
This is where the second common misconception breaks: not all public spending can be cut each year. Legislators can reopen a discretionary programme at the next budget; they cannot simply decline to pay a pension already earned or a coupon already promised.
Analysis: Debt service is a first claim because the alternative is default, and default is expensive in ways that outlast the government that risks it: lost market access, a sharp spike in future borrowing costs, and damage to domestic banks holding the bonds. Interest is not a program with a constituency; it is a bill with a deadline.
The arithmetic is unforgiving. Take an illustrative economy with output of $1,000 and a debt stock of $600, a debt-to-GDP ratio of 60 percent. At an effective interest rate of 4 percent, debt service is , or 2.4 percent of output. Suppose tax revenue is $300 and mandatory transfers are $156, leaving for discretionary programs. Now revenue disappoints by 10 percent, to $270, while debt service and mandatory spending stay put. Discretionary room becomes — a 25 percent cut in the flexible part of the budget caused by a 10 percent revenue shortfall. If the effective rate on the same debt rises from 4 to 5 percent, interest becomes , and discretionary room falls again, to $84.
flowchart TD
A["Tax revenue: direct and indirect"] --> B["Debt service paid first"]
B --> C["Mandatory transfers paid"]
C --> D["Discretionary spending: the residual"]
D -->|"protected instead of cut"| E["Borrowing"]
E -->|"larger debt stock"| F["Higher debt service next year"]
F --> B
The loop is the point: protecting today's discretionary spending with borrowing raises tomorrow's first claim, shrinking the residual again.
This is why budget failure is about consent as much as arithmetic. The size of the gap is set by the numbers; where it lands is decided by veto players who each defend a different line. Every group treats the revenue pool as a shared resource and its own claim as sacred — the common-pool problem in its purest form. A government can pass a budget that is arithmetically identical to one that failed a month earlier, because what changed was the political allocation, not the total.
Conclusion
A public budget must satisfy an accounting constraint and survive a political one. Direct taxes are visible and contested; indirect taxes are quieter but no less real a cost to households. Debt service and mandatory commitments are prior claims, so discretionary spending absorbs the shock — which is why the flexible part of a budget can shrink far faster than revenue does. Past borrowing is not a settled historical fact; it is a standing claim on present revenue, priced by whoever holds the bonds. That is where the next chapter picks up: borrowing is not a free third option, because the market sets a price for it, and that price is the political ceiling made visible.
1.3: Why Budgets Fail Even When Everyone Agrees on Arithmetic: Veto Players and the Common Pool
A finance minister stands before a chart. The deficit projection is precise to the decimal. Every party in the room says the arithmetic is correct. Then the budget dies. The puzzle is not numeracy. A public budget must satisfy an accounting identity: total spending equals revenues plus net borrowing. But a budget must also satisfy a coalition. The previous section showed how tax and spending claims are made; here those claims become competing vetoes. Budget failure is often a consent problem, not mere irrationality.
Veto players
A veto player is any actor whose agreement is required for a budget change. Institutional veto players come from constitutional rules: an executive veto, a second chamber, a court, or a coalition partner with a blocking position. Partisan veto players come from politics: parties whose votes are needed to pass a package. Tsebelis formalized the point: the more veto players there are, and the farther apart their preferred policies, the smaller the set of changes that can beat the status quo #(Tsebelis, 2002, Veto Players: How Political Institutions Work). In plain terms, a budget can be arithmetically correct and still have no majority for any distribution of the burden.
This is not a story of stupid politicians. It is a story of credible threats. A coalition partner can block a tax increase that hurts its voters; an opposition chamber can reject a spending cut; a finance minister can refuse to sign off on a deficit path. Each has a reason. Alesina and Drazen call the result a war of attrition: when adjustment is unavoidable, each faction waits, hoping another absorbs the cost first #(Alesina and Drazen, 1991, Why Are Stabilizations Delayed?). Delay is rational for each player, ruinous for the collective.
flowchart TD
A["Shared tax pool"] --> B["Competing spending claims"]
B --> C["Veto player 1"]
B --> D["Veto player 2"]
B --> E["Veto player 3"]
C --> F{"All agree on distribution?"}
D --> F
E --> F
F -->|"Yes"| G["Budget passes"]
F -->|"No"| H["Delay, stopgap or partial package"]
H --> I["Rollover risk and higher debt service"]
The diagram is not a prediction; it is a map of the consent gate. A budget can pass procedurally through a stopgap or a decree, but that is not the same as durable political agreement. The paradox of institutional checks is that they protect minority interests and constrain arbitrary action, yet during a shock they can also make fiscal adjustment slower. The literature finds that multi-party coalition governments, with multiple partisan veto players, often experience longer delays in consolidation after macroeconomic shocks #(Roubini and Sachs, 1989, Political and Economic Determinants of Budget Deficits in the Industrial Democracies). Franzese finds similar persistence when institutional veto players multiply #(Franzese, 2002, Macroeconomic Policies of Developed Democracies).
The second mechanism is the common-pool problem. Public revenue is a shared pool. Each spending ministry or coalition partner can claim it; the tax cost is spread across everyone. Weingast, Shepsle and Johnsen show that decentralized bargaining over targeted projects can make spending rise with the number of decision-makers, because each internalizes the benefit to its district but only a fraction of the cost #(Weingast, Shepsle, and Johnsen, 1981, The Political Economy of Benefits and Costs: A Neoclassical Approach to Distributive Politics). Their logic is simple: if a project’s benefit is concentrated and its cost is divided across many taxpayers, the claim looks cheap to the claimant and expensive only in aggregate. Olson’s logic reinforces it: concentrated beneficiaries organize; diffuse taxpayers do not #(Olson, 1965, The Logic of Collective Action). The pool is overfished, not because everyone is irrational, but because the returns to claiming are concentrated and the costs are diluted.
Analysis: The common-pool problem explains why broad agreement on the deficit number does not produce agreement on the deficit path. The number is a residual; the fight is over who pays. A budget rule can tighten the arithmetic, but enforcement runs through veto players. Hallerberg, Strauch and von Hagen show that fiscal governance works through institutions, not magic numbers #(Hallerberg, Strauch, and von Hagen, 2009, Fiscal Governance in Europe). Centralizing agenda power in a strong finance minister can help internalize the cost #(von Hagen and Harden, 1995, Budget Processes and Commitment to Fiscal Discipline). But centralization is itself a political choice, and it can be reversed.
Persistent veto bargaining also travels to markets. Poterba and von Hagen link fiscal stalemate to weaker budget performance and higher borrowing costs: when institutions look unable to consolidate in time, sovereign bond markets may widen spreads over benchmark rates and raise rollover risk #(Poterba and von Hagen, 1999, Fiscal Institutions and Fiscal Performance). That link is not automatic. The threshold at which markets penalize gridlock is non-linear and depends on monetary sovereignty and central-bank credibility. The safe conclusion is directional: consent failure raises the price of waiting.
Conclusion
What must balance? The accounting identity. What must be agreed? The distribution of costs. Budget failure is often a failure of consent, not a failure of arithmetic. The arithmetic sets the size of the adjustment; veto players set whether and when it happens. Organized groups defend concentrated benefits, while diffuse taxpayers absorb the costs, so even a technically passed budget may remain politically fragile. This is the political backdrop for the market repricing examined in Chapter 2 and the institutional vocabulary for the three-legged stool in 3.1. A budget is both a constraint and a coalition document; reading it as only one or the other misses why governments fail even when everyone can add.
Chapter 2: The Price of the Promise—How Markets Read Fiscal Choices
2.1: A Yield Is a Price, Not a Bill: Bonds, Spreads and the Signal That Matters
On a Tuesday morning, a government funding desk sells ten-year bonds at a yield of 5%. By lunchtime the headline reads: "Borrowing costs hit 5%." That afternoon a backbencher asks the obvious question — does the state now pay 5% on everything it owes? The budget papers say the interest bill is nearer 2% of GDP. Both numbers can be true at once, because they describe different objects. One is a price on the next bond; the other is an average across a stock built up over decades. Chapter 1 treated borrowing as the residual left when revenue and spending cannot easily adjust. This subchapter is about how the market prices that residual — and why the price is not the same thing as the bill.
Sovereign bond yield
A bond is a promise: fixed coupons each year, principal at maturity. The yield is the single discount rate that makes the present value of those promised payments equal today's market price #(Fabozzi, 2007, Fixed Income Analysis). Change the price and the yield changes with it — the promise itself is untouched. Think of a shop selling a fixed, unchangeable voucher. If the voucher's price tag drops, the return to whoever buys it rises. That is the whole of the inverse relationship: price down, yield up; price up, yield down.
The coupon and the yield are therefore different animals. A bond issued years ago paying a 2% coupon still pays 2%, forever. If it now trades at 80 cents on the euro, a new buyer collects that 2% coupon plus a capital gain at maturity — an annualised return of roughly 5%. The issuer, meanwhile, is still paying 2% on that particular bond.
This is why a headline yield is a marginal price. It is what the market charges for new money today — a few billion at auction, plus whatever matures and must be refinanced. The effective interest rate is the average cost of the whole debt stock, weighted by everything issued in the past.
| Dimension | Marginal yield | Effective interest rate |
|---|---|---|
| What it prices | The next bond issued or rolled over | The entire outstanding stock |
| How it is set | Daily market trading and auctions | History of issuance and coupons |
| How fast it moves | Minutes | Months and years, one maturity at a time |
| Budget effect today | None | Drives the debt-service line |
Here is the arithmetic, with clearly illustrative numbers. A country carries 1,000 units of debt, equal to 100% of GDP. Nine hundred sit in older bonds averaging 2%; 100 mature this year and are refinanced at 5%.
- Interest on the old stock: .
- Interest on the refinanced slice: .
- Total interest: , an effective rate of of the stock.
The effective rate rose from 2.0% to 2.3% — three-tenths of a point of GDP, not three points. And if average maturity is seven years, only about one-seventh of the stock reprices annually, so full pass-through takes years. With two-year average maturity, the same 5% bites far sooner. That maturity profile is precisely what feeds the effective-rate term in the debt-snowball equation we will meet in 2.2.
Analysis: A yield is a thermometer reading taken on the marginal bond, not an invoice for the stock. Confusing the two exaggerates today's fiscal pain and quietly understates tomorrow's.
Three errors follow from that confusion. First, treating a 5% yield as a 5% charge on all debt overstates the immediate interest bill by an order of magnitude. Second, imagining that rising yields help a government — they do not. Higher yields are good for the buyer, not the issuer; the sovereign only ever pays more, and the higher price also raises the cost of every future rollover. Third, blaming every yield move on national policy: much of it is global. When the US Treasury or German Bund yield shifts, every sovereign's yield tends to shift with it.
That is why the diagnostic number is the spread: the country's yield minus a safe benchmark of the same maturity #(Codogno, Favero, and Missale, 2003, Yield Spreads on EMU Government Bonds). A yield bundles together the expected path of policy rates, an inflation and duration premium, and a country premium. The first two are shared weather; the third is the country's own climate.
flowchart LR
A["Spread over benchmark widens"] -->|"new bonds priced higher"| B["Marginal yield rises"]
B -->|"only maturing bonds refinance"| C["Effective interest rate creeps up"]
C -->|"debt service grows"| D["Fiscal space shrinks"]
Italy in 2011 is a compact illustration rather than a full case. Its ten-year spread over the German Bund widened past 500 basis points as political credibility came into question, while its headline deficit was modest by the standards of the period. The market repriced the marginal bond sharply; the effective interest cost on the existing stock climbed far more slowly #(Reinhart and Rogoff, 2009, This Time Is Different: Eight Centuries of Financial Folly). Japan supplies the mirror image: a debt ratio well above Italy's with yields near zero, reflecting domestic ownership, institutional structure and central-bank presence #(Hoshi and Ito, 2014, The Japanese Economy). Similar debt levels, different outcomes — because the spread prices institutions and expectations, not just arithmetic.
Analysis: A spread is a signal that can also become a self-fulfilling prophecy. Sovereign debt markets can host more than one equilibrium, and a credible backstop can move the price without any change in the debt itself #(Calvo, 1988, Servicing the Public Debt: The Role of Expectations). Read spreads with humility: they diagnose, but they also participate.
Conclusion
Three distinctions do most of the work. A yield is a price, computed from the bond's market value, and it moves inversely with that value. It is a marginal price on new borrowing, not an average bill on the existing stock. And it is most informative when read as a spread over a safe benchmark, because only the spread isolates what is genuinely country-specific from what is happening to interest rates everywhere. Hold on to the effective-rate mechanism: the marginal yield is the tap, the effective rate is the reservoir, and the debt stock is the water already in it. That slow, one-maturity-at-a-time transmission is what turns a market signal into a fiscal fact — and it is exactly where the snowball dynamics of the next subchapter begin.
2.2: The Snowball Equation Without the Panic: r − g, Primary Balance and the Limits of Forecasting
Suppose a finance ministry wakes to news that the yield on its ten-year sovereign bond has jumped from 2% to 4%. The headline writes itself: borrowing costs have doubled. The budget desk knows better. Most of the debt was issued years ago at older coupons. The higher yield applies only to new bonds and to maturing tranches as they roll over. Yet the market move still matters, because eventually the marginal yield becomes the effective interest rate—the weighted average cost of the whole debt stock, the idea introduced in 2.1—and that average is what feeds debt service.
r − g debt dynamics
Start with the difference between a price signal and a bill. An investor buying a sovereign bond today earns the market yield; the government pays coupons and principal. The spread over a benchmark strips out global inflation and safe-asset moves, so it is the cleaner read on sovereign-specific risk. But neither the headline yield nor the spread enters the public budget immediately. The effective interest rate rises only as old debt matures. A long weighted average maturity is a shock absorber; a short one is rollover risk.
The budget identity makes the arithmetic visible. Let be the debt stock and the primary balance—revenue minus non-interest spending, with a primary deficit negative. Then:
Divide by GDP, which grows at , and the debt-to-GDP ratio follows:
After subtracting inflation, and using the small-step approximation, the familiar snowball equation appears:
Here is the real effective interest rate, is real growth, and is the primary balance as a share of GDP. This is an identity, not a forecast.
Take a simple example. Suppose debt is 100% of GDP, , , and the primary deficit is 0.5% of GDP. The snowball term is of GDP. Subtracting a primary deficit of gives a debt-ratio increase of about 1.5% of GDP. To stabilize debt with , the government needs a primary surplus of roughly 1% of GDP, since .
When , the snowball is a headwind: past debt compounds faster than the economy grows unless a primary surplus offsets it. When , the wind reverses; an economy can carry a primary deficit without the ratio diverging #(Blanchard, 2019, Public Debt and Low Interest Rates). This is why similar debt levels can produce different outcomes. The Japan puzzle—very high debt ratios alongside very low yields—reflects domestic investors, central-bank backing and a favorable regime, not proof that debt ratios never matter.
Now separate the two balances. The headline balance is the primary balance minus interest payments. If a government runs a primary surplus of 2% of GDP but pays interest of 4%, its headline deficit is 2%. Confusing the two is like judging a household’s current spending by including interest on an old mortgage.
| Balance | What it includes | What it reveals |
|---|---|---|
| Primary balance | Revenue minus non-interest spending | Current fiscal effort |
| Headline balance | Primary balance minus interest | Total borrowing need |
The primary balance is the policy lever in the snowball equation; the headline balance is the bottom line the bond market finances.
The next distinction prepares the three-legged diagnostic in 3.1. A structural deficit adjusts for the business cycle and one-off factors; a cyclical deficit comes from automatic stabilizers as tax revenue falls and spending rises in a downturn. A recession can widen the deficit without a permanent deterioration in fiscal effort. Misreading a cyclical shortfall as structural invites austerity that may cut the primary deficit while also cutting —and a weaker denominator can push the debt ratio up, not down #(DeLong and Summers, 2012, Fiscal Policy in a Depressed Economy).
Analysis: The debt equation is a guide to assumptions, not a prophecy. Forecasts must show their , , primary-balance and maturity assumptions; stress tests should let spike while falls. Forecasts define fiscal space, but they cannot abolish uncertainty.
Conclusion
Return to the yield puzzle. The headline yield is a signal; the effective interest rate is the transmission channel; the primary balance is the fiscal effort; is the weather. Watch the spread, the maturity profile and the structural-versus-cyclical split. Then treat the debt path as a visible set of assumptions—useful for navigation, dangerous as prophecy. The snowball equation does not tell us precisely what will happen. It tells us what would have to be true for debt to stabilize, and that is already a great deal.
2.3: The Japan Puzzle and the Greece Warning: Four Ways to Read the Same Debt Number
Imagine two governments that each owe more than their economies produce in a year. One borrows for a decade at less than 1%; the other pays double digits. That was roughly Japan and Greece at the height of the euro crisis. Japan’s gross debt was above 200% of GDP and its ten-year yield stayed near zero; Greece’s debt was smaller and its borrowing cost exploded. The debt number alone cannot explain the gap. The answer lies in four lenses: who owns the debt, what currency it is in, when it must be repaid, and who stands behind it.
Debt ownership and investor base
Ownership is the first lens because a bond is a promise to a particular holder. Japan is the cleanest illustration: its enormous gross debt is overwhelmingly held at home, by banks, insurers, pension funds and the Bank of Japan. That home bias is sustained by regulation, institutional habit and the search for safe yen assets #(Horioka et al., 2014, The High and Durable Domestic Ownership of Japanese Government Bonds). A domestic base does not make debt disappear. It changes the speed of trouble. Domestic holders cannot easily flee into another currency, and the central bank can act as buyer of last resort. The cost is a bank-sovereign connection: when a government’s bonds sit on domestic banks’ books, a fiscal shock can become a banking shock.
Greece before 2010 was closer to the mirror image. Its debt was in euros, a currency the Greek state did not control in the way Japan or the United Kingdom controls its own, and a large share was held by foreign private creditors. When confidence cracked, those investors could sell or refuse to roll over, and there was no national central bank able to promise unlimited euro support #(Lane, 2012, The European Sovereign Debt Crisis). The official rescues that followed changed ownership again, tying Greece to conditionality and euro-area politics. France shares Greece’s currency constraint but not its recent history: it has a deep domestic institutional base and relatively long maturities, so political uncertainty tends to show up first as a wider spread over German Bunds rather than as an immediate refinancing crisis. That is why France is not a pure arithmetic story. The United Kingdom sits on the other side: it borrows in sterling and has its own central bank, so its large debt has not produced Greek-style yields. But the 2022 gilt episode showed that a backstop is a conditional cushion, not a guarantee against a sharp marginal repricing.
Currency, maturity and backstop are the other three lenses. Currency determines whether a government can always find the nominal money to pay domestic-currency debt, though not without inflation risk. Maturity determines how often old debt is repriced: short bills transmit market pressure quickly; long bonds delay it. The backstop determines who can bridge a rollover. Japan and the UK have national central banks; France and Greece rely on the European Central Bank, a collective institution with conditions and political limits. The ECB can calm markets, but it does not turn a euro member into a yen or sterling issuer #(De Grauwe, 2012, The Governance of a Fragile Eurozone).
flowchart LR
A["Political uncertainty"] --> B["Marginal bond yield rises"]
B --> C{"Domestic owners and national backstop?"}
C -->|"Yes"| D["Spread widens but rollover is cushioned"]
C -->|"No"| E["Refinancing stress rises quickly"]
D --> F["Average effective interest rate adjusts slowly"]
E --> F
F --> G["Debt momentum worsens if average rate exceeds growth"]
Those lenses explain why the same debt ratio can carry different market prices, but they do not explain which price to watch. A ten-year yield is a marginal price: what a government pays to borrow new money today, or what existing bonds trade at. It is not the average interest rate on the whole debt stock. Most outstanding bonds were issued years ago at older coupons, so a jump in the ten-year yield feeds into the budget only as debt matures and is refinanced #(Abbas et al., 2019, Sovereign Debt: A Guide for Economists and Practitioners). A political shock can therefore raise headline yields without an immediate jump in debt service.
| Measure | What it tells you | What it does not tell you |
|---|---|---|
| Marginal yield | Price of new borrowing at a given maturity | The whole stock’s interest bill |
| Average effective interest rate | Actual average cost of outstanding debt | Today’s market pressure |
| Spread over benchmark | Country-specific risk and liquidity premium | The absolute level of global rates |
| Debt-to-GDP | Debt stock relative to annual output | Ownership, currency, maturity or backstop |
The spread is more diagnostic than the absolute yield. If global inflation lifts all bond yields, a country’s borrowing cost can rise without its credibility changing; the spread over a benchmark strips much of that common move out. Within the euro area, the spread over German Bunds is the standard read on country risk and liquidity #(Codogno et al., 2003, Yield Spreads on EMU Government Bonds). But even a spread is not a verdict: it can widen because of political uncertainty, because investors demand more compensation for a less liquid bond, or because they expect weaker growth.
Debt-to-GDP is the final limit, and the most misused. It is a stock relative to annual output, not a price and not a solvency test. Japan’s gross ratio is far above Greece’s pre-crisis ratio, yet Japan’s yields have been far lower. Net debt, which subtracts liquid assets, can look different; ownership, currency, maturity and backstop can change the meaning again. There is no universal threshold at which a debt ratio triggers a crisis. Markets and institutions set the price, and the price can move before the ratio does.
Analysis: High debt does not always produce high yields; it produces vulnerability that institutions either absorb or amplify. A debt snowball accelerates when the average effective interest rate on the stock exceeds nominal growth—when —and the primary balance cannot offset it. Marginal yields matter because they eventually become average costs, but only through rollover. A country with long maturities and a credible backstop can carry a high ratio for years; a country with short maturities, foreign-currency exposure and no backstop can be pushed into crisis at a lower ratio. Greece is a warning about one dangerous combination, not a universal template.
Conclusion
Read the debt number, then read the four lenses. Ownership tells you who can leave; currency tells you what the government can create; maturity tells you how fast market prices reach the budget; the backstop tells you who can stop a run. Japan shows that high debt can coexist with low yields when a domestic base and a national central bank absorb the pressure. Greece shows what happens when external creditors, a shared currency and no national backstop combine. Britain shows that a backstop still has to be credible, and France shows that political risk can widen a spread long before it becomes an arithmetic emergency. This is the comparative evidence the three-legged stool in 3.1 needs: debt level, market price and institutional credibility are separate legs, and no single debt ratio can stand in for all three.
Chapter 3: Choosing Under Constraint—The Three-Legged Stool and the Policy Menu
3.1: The Three-Legged Stool: Arithmetic, Institutions and the Backstop Question
In the summer of 2012, two governments sat at opposite ends of the same bond market. Japan carried gross debt above 250 percent of GDP and borrowed at rates near zero. Spain, with a debt ratio a fraction of Japan's, paid over 7 percent on ten-year money and watched its refinancing calendar with dread. The previous chapter established that debt levels alone cannot explain borrowing costs. So what does? The answer is not a single number but a diagnostic with three legs, and a stool that stands or falls on the weakest of them.
Three-legged fiscal sustainability diagnostic
Leg one: arithmetic. The debt ratio moves according to a budget identity you can write on an envelope: , where is debt-to-GDP, the real interest rate actually paid on the stock, real growth, and the primary balance (revenue minus non-interest spending, as a share of GDP). Take a country with debt at 100 percent of GDP, . If and , the snowball term is : growth alone shrinks the debt ratio by 4 points of GDP a year, so a primary deficit of up to 4% still stabilises debt. Now flip the assumptions: falls to 1% and the marginal yield pushes the effective rate to 4%. The term becomes , and the same country must run a primary surplus of 3% of GDP merely to stand still. That swing — from comfortable deficit to punishing surplus, with no change in borrowing behaviour — is why the arithmetic leg is a test of sensitivity, not a verdict #(Domar, 1944, The 'Burden of the Debt' and the National Income; Blanchard, 2019, Public Debt and Low Interest Rates).
Analysis: Note what the equation uses: is the effective rate on the whole stock, not today's marginal yield. A government that borrowed long and cheap has time; one that must refinance a third of its debt within a year has almost none. Maturity structure is where arithmetic meets the market signal from Subchapter 2.1.
Leg two: institutions. A debt path is a promise about future budgets, and promises are only as good as the machinery that delivers them. Two mechanisms matter. The common-pool problem means each spending ministry or regional government captures the full benefit of its own outlay while sharing the cost with everyone else, so deficits drift upward absent a central counterweight #(Weingast, Shepsle, and Johnsen, 1981, The Political Economy of Benefits and Costs: A Neoclassical Approach to Distributive Politics). Veto players — coalition partners, upper chambers, constitutional courts — determine how many doors a consolidation must pass through, and more doors mean slower, more expensive adjustment #(Tsebelis, 2002, Veto Players: How Political Institutions Work). This is the leg that separates Japan's high debt from Spain's expensive debt: roughly similar arithmetic pressure, very different capacity to legislate a primary balance.
Leg three: the backstop. A sovereign borrowing in its own currency, through a central bank that can buy its bonds, faces a different market than one borrowing in a currency it does not control. The backstop works by eliminating the bad equilibrium: if investors believe a run is possible, they run, and the fear becomes self-fulfilling #(Calvo, 1988, Servicing the Public Debt: The Role of Expectations; De Grauwe, 2012, The Governance of a Fragile Eurozone). The European Central Bank's 2012 pledge to do "whatever it takes" collapsed periphery spreads without spending much money, because it removed the run, not the debt.
But here is the conditional core, and the most common misconception in this entire subject: a backstop that exists is not a backstop that will be used. An unconditional guarantee destroys the incentive to adjust — if rescue is certain, delay is rational for every veto player — so central banks attach conditions before activating purchases. The OMT programme was never triggered precisely because its conditionality was credible. Analysis: Read backstop announcements as contingent contracts, not insurance policies. The question is never "does the central bank have the tool?" but "what will it demand first, and will the government be able to comply?"
| Leg | Question it asks | Failure looks like |
|---|---|---|
| Arithmetic | Does the debt path stabilise under plausible and ? | Interest costs crowd out everything else |
| Institutions | Can a correction actually be legislated? | Repeated promises, missed targets, rising spreads |
| Backstop | Is the debt in a currency the sovereign can backstop, and on what terms? | A liquidity scare becomes a solvency crisis |
The third misconception is subtler: a government promising to adjust is not evidence of anything, because of time inconsistency. A finance minister who announces a future consolidation faces a temptation to renege once wages, contracts and elections have adjusted to the announcement; markets anticipate this and price it in, so discretion produces higher borrowing costs than a rule would #(Kydland and Prescott, 1977, Rules Rather than Discretion: The Inconsistency of Optimal Plans). Commitment devices — medium-term expenditure frameworks, independent fiscal councils, debt brakes — exist to make the promise costly to break. They are not guarantees of virtue; they are ways of raising the price of backsliding. And they have a cost: rigid rules can smother the automatic stabilisers exactly when a downturn needs them.
Analysis: No leg is sufficient. Strong arithmetic with no institutional delivery fails at the first recession; strong institutions with no backstop can be flattened by a self-fulfilling panic; a backstop without arithmetic or credibility merely converts a default into an inflation.
Conclusion
The three-legged diagnostic turns an argument about debt levels into three answerable questions: does the arithmetic work under plausible rates and growth, can the political system deliver the correction, and is there a conditional backstop standing behind the debt? When a fiscal constraint becomes binding, it is almost never because one number crossed a threshold — it is because two or three legs gave way at once. The practical payoff is a disciplined way to judge any fiscal proposal: ask what it assumes about , what institution will enforce it after the election, and what happens if the market's mood shifts before the backstop is invoked. Those same three questions organise the policy menu in Subchapter 3.2, where the tools of consolidation — spending, taxing, borrowing, and the institutional packaging that makes any of them credible — meet their trade-offs.
3.2: From Diagnosis to Choice: A Trade-off Menu for Fiscal Decisions
The finance ministry opens its morning meeting with one number: the sovereign’s marginal yield has risen 220 basis points over benchmark, and a fifth of the debt stock matures within a year. The three-legged diagnostic says the market leg is binding. But that diagnosis does not choose for the government. It can cut spending, raise taxes, invest for growth, rebuild budget institutions, or re-engineer the debt profile. Each option is a different bet on credibility, timing and who bears the cost.
Fiscal consolidation versus growth strategy
The arithmetic is the first constraint. The debt-to-GDP ratio moves according to the budget identity , where is debt-to-GDP, is the real effective interest rate, is real growth and is the primary balance. Start with debt at of GDP, , and a primary balance of zero. The debt ratio rises by percentage points a year. Stabilising it requires a primary surplus of of GDP. That is arithmetic, not ideology.
But consolidation is not a dial. If the fiscal multiplier is 1.5, a consolidation of of GDP can cut output by , so the debt ratio may rise before it falls. #(DeLong & Summers, 2012, Fiscal Policy in a Depressed Economy) That is why timing matters: front-loaded adjustment can reassure markets but deepen a downturn; back-loaded adjustment can protect growth but invite doubts about commitment. The target should be the structural deficit, not the headline deficit, because automatic stabilizers move the cyclical deficit with the cycle.
Design matters too. Spending-based consolidations have often been less contractionary than tax-based ones, partly because tax increases can dampen investment and labour supply. #(Alesina & Ardagna, 2010, Large Changes in Fiscal Policy: Taxes versus Spending) But the distributional cost is not the same. Spending cuts often land on transfers, public wages and maintenance; tax rises can land on consumption and small firms. Analysis: Credibility is not the same as severity. A credible plan is one markets and voters expect to survive, not the largest cut announced.
Growth strategy is the second lever. It aims to raise through public investment, skills, competition and structural reform. If a project returns and the sovereign’s marginal yield is , debt-financed spending worsens the debt ratio unless spillovers are large. #(Blanchard, 2019, Public Debt and Low Interest Rates) Growth is not a panacea. Japan’s high debt and low yields reflect a central-bank backstop, home bias and deflation—not a portable template. In the euro area periphery after 2010, countries without a national central-bank backstop faced a different sequence: consolidation first, output losses second. #(De Grauwe, 2012, The Governance of a Fragile Eurozone)
Institutional reform is slower and less visible. Judicial reform, procurement rules and a credible medium-term budget can raise growth and improve credibility, but they pay over years. Veto players can block or reverse them; the common-pool problem biases fragmented budget processes toward deficits. #(Tsebelis, 2002, Veto Players: How Political Institutions Work) Debt management can buy time—lengthening maturity, smoothing the effective interest rate, reducing rollover risk—but it does not reduce the debt stock. Debt mutualisation or inflation can lower the real burden, but both redistribute: mutualisation shifts risk across countries and can create moral hazard; inflation hurts savers and wage earners and strains the central bank.
When judging any proposal, ask three questions:
- Credibility: who will implement it, and what makes reversal costly?
- Time horizon: does it pay before the next rollover wall?
- Distribution: who bears the first, visible cost—and who bears the deferred one?
The menu is therefore not a ranking. If the market leg is binding—wide spreads, short maturity, weak auctions—consolidation is the entry ticket, but its design should protect public investment and the poor. If the arithmetic is binding but markets are patient, growth and institutional reform can be sequenced first. If institutions are the weak leg, budget-process reform may be the highest-return option. Each choice has a cost; none is painless.
Conclusion
The three-legged diagnostic tells a government which constraint bites. It does not tell it whom to ask to pay. A defensible fiscal choice names the binding constraint, sets a timeline that matches the rollover calendar, and states the distributional trade-off. Consolidation can restore credibility but may deepen a downturn; growth can raise the denominator but may arrive too late; institutional reform can improve both credibility and growth but is slow and uncertain; debt management can smooth the path but cannot erase a debt stock. Fiscal space is not a fixed number; it is the room between current policy and the point where the market leg becomes a wall. The honest answer is conditional—and it should be reviewed as the market signal and the budget outturn change.
Conclusion
A budget is both arithmetic and a political settlement: deficits add to debt, while taxes and spending determine who bears the adjustment and when. Because debt service and mandatory commitments are hard to cut, higher yields can squeeze flexible public services; direct and indirect taxes distribute the burden differently. Markets judge not just debt levels but growth, political consent, and credible backstops—factors that shape borrowing costs and credit ratings. Japan, Greece, Britain, and France show why context matters. If yields rise, ask: what must change, who pays, and can the plan survive a downturn? Weigh austerity’s credibility against stimulus’s support for growth, then test each choice against the budget’s arithmetic, politics, and institutions.
⟬
- Which statement best describes the public budget identity? A. Revenue minus expenditure equals the deficit, which equals new borrowing.© B. The deficit is the sum of revenue and expenditure. C. New borrowing is the difference between expenditure and the debt stock. D. The primary balance includes interest payments on existing debt. The budget identity states that revenue minus expenditure equals the deficit, and the deficit equals net new borrowing. The primary balance excludes interest, and the debt stock is a stock, not a flow.
- Which of the following is a direct tax rather than an indirect tax? A. Value-added tax B. Sales tax C. Personal income tax© D. Excise duty Direct taxes fall on income, wealth, or profit. Personal income tax is a direct tax, while VAT, sales tax, and excise duty are indirect taxes on transactions.
- What is a veto player in the context of budget negotiations? A. Any actor whose agreement is required for a budget change.© B. A politician who votes against the budget but cannot block it. C. A market participant who sets the bond yield. D. An economist who forecasts the deficit. A veto player is any actor whose agreement is necessary for a budget change, such as coalition partners, second chambers, or courts. This institutional and partisan dynamic can prevent budget passage even when the arithmetic is correct.
- Why does a 5% marginal yield on new government bonds not immediately raise the government's average interest bill to 5%? A. Because the effective interest rate is the average cost of the whole debt stock, and most debt is issued at older rates and matures gradually.© B. Because bond yields and interest bills are unrelated. C. Because governments do not pay interest on newly issued bonds. D. Because the central bank always covers the difference. The marginal yield applies only to new borrowing. The effective interest rate is the weighted average cost of the entire outstanding debt, which adjusts slowly as old debt matures and is refinanced.
- In the debt snowball equation, what does the term (r - g) represent? A. The difference between the real effective interest rate and real GDP growth.© B. The difference between nominal interest rate and inflation. C. The difference between the primary balance and the headline balance. D. The difference between debt and GDP. In the equation Δb ≈ (r - g)b - pb, r is the real effective interest rate and g is real growth. The difference (r - g) determines whether debt compounds faster than the economy grows, requiring a primary surplus to stabilize debt if positive.
- According to the three-legged stool diagnostic, what is the third leg that matters for fiscal sustainability? A. The currency backstop, including whether debt is in the sovereign's own currency and whether a credible central bank or lender-of-last-resort facility exists.© B. The level of government spending as a share of GDP. C. The size of the primary deficit relative to GDP. D. The maturity structure of the debt. The three-legged stool comprises arithmetic (r - g and debt path), institutions (capacity to deliver primary balance), and the currency backstop. The backstop leg addresses whether the sovereign can backstop its debt, often through a central bank, which affects market confidence and run risk. ⟭