France’s Debt: A Reminder of Responsibility

France is entering its 2027 presidential campaign with public debt at a level that would have seemed extraordinary a generation ago. Jean-Luc Mélenchon, the declared presidential candidate of La France insoumise (LFI), has revived a long-standing proposal to cancel or neutralise the part of French public debt held within the Eurosystem. In recent posts he has also presented public debt as a deliberate strategy used to destroy social gains and enrich privileged groups. Éric Coquerel, the LFI chairman of the National Assembly’s Finance Committee, has defended a related argument for years, notably in his book Lâchez-nous la dette ! (“Get the Debt Off Our Backs”). [S1][S2][S3][S22]

For readers outside France, the political context matters. France has one of the largest welfare states in Europe, very high public expenditure, a long tradition of central government intervention and, since 1975, an uninterrupted series of annual public deficits. The debate is therefore not simply “austerity versus generosity”. It is about whether a country can permanently spend more than it raises, refinance the difference through markets, and later describe the resulting stock of debt mainly as an external imposition.

My disagreement with Mélenchon and Coquerel is not that monetary institutions are beyond criticism. The ECB’s mandate, Article 123 of the EU treaty, the distributional effects of monetary policy and the economic merits of public investment are all legitimate subjects for debate. Nor is debt inherently bad. My objection is historical and accounting-based: France’s debt did not first appear because creditors forced France to borrow. It accumulated because successive governments, from the left, right and centre, repeatedly voted for expenditure above revenue and financed the gap.

Debt is not bad by nature

The IMF itself makes an important distinction: debt is not good or bad in itself. It can smooth recessions, finance emergency responses and spread the cost of long-lived investment across the generations that benefit from it. A railway, power grid, hospital, defence programme or climate adaptation project may rationally be financed over decades instead of entirely by one year’s taxpayers. [S21]

Olivier Blanchard, former IMF chief economist, has also shown why the interest-rate/growth relationship matters. If safe interest rates remain below nominal economic growth for long periods, the fiscal cost of debt can be lower than intuition suggests. That is a reason to reject simplistic anti-debt rhetoric. It is not, however, a reason to ignore permanent primary imbalances. [S20]

France’s problem is that borrowing ceased to be only a tool for exceptional shocks and productive investment. According to the historical series reproduced by the National Assembly, France last recorded a positive public balance in 1974, at only 0.1% of GDP. From 1975 onward, every year in that series is in deficit. [S5]

That half-century matters more than any slogan. Each annual deficit is a flow. Repeated flows become a stock. Interest then compounds the burden, especially when old low-rate bonds have to be refinanced at higher rates.

France’s debt has reached €3.536 trillion

At the end of the first quarter of 2026, France’s Maastricht public debt stood at €3,536.1 billion, or 117.5% of GDP. It had been 115.7% of GDP at the end of 2025. The quarterly increase was €75.6 billion, although Insee correctly warns that a quarterly change in debt is not identical to the deficit because cash and financial-asset movements also matter. [S4]

The historical contrast is stark. In 1980 debt was about 20.7% of GDP. France has moved from a debt stock equivalent to roughly one fifth of annual output to a stock exceeding one year of output. [S6]

The cost is now becoming as politically important as the stock. France’s Court of Auditors estimated in February 2026 that interest expenditure would approach €74 billion in 2026 and could exceed €100 billion by 2029 if the current dynamics persisted. The OECD likewise noted that debt-service costs rose from 1.3% of GDP in 2020 to 2.1% in 2025. [S15][S16]

This is why debt sustainability is not an abstract concern imported from financial markets. Every euro devoted to servicing past commitments is an euro that cannot simultaneously finance schools, hospitals, defence, justice, infrastructure, research or the energy transition.

The French left is a co-author of this history, not an outside victim

It would be false to attribute fifty years of deficits to the left alone. Conservative governments, centrist governments and the Macron era all contributed. The 2008 financial crisis and the Covid pandemic also generated large, understandable emergency borrowing. A serious account has to say that clearly.

But collective responsibility makes today’s rhetoric more difficult, not less. When François Mitterrand entered the Élysée in 1981, public debt was about €110 billion in converted values and 22% of GDP. By 1986 it was around €249 billion and 31.1% of GDP. In 1988 it was roughly €303 billion and 33.3% of GDP; by 1993 about €515 billion and 46%; and in 1995 about €664 billion and 55.5% of GDP. These calendar-year figures span governments and cohabitations, so they cannot be assigned mechanically to one party. They nonetheless show that the left was deeply involved in the period during which the modern French debt stock expanded. [S6]

The same honesty must be applied in the other direction. Under Socialist prime minister Lionel Jospin, the debt ratio did not explode and in favourable economic conditions fell relative to GDP for part of the period. Debt is not a genetic property of one political label.

Under Socialist president François Hollande, however, Maastricht debt rose from about €1,892.5 billion and 90.6% of GDP at end-2012 to €2,258.7 billion and 98.4% at end-2017. That is roughly €366 billion of additional debt over those calendar endpoints. [S7]

The conclusion is not “the left caused everything”. The conclusion is that nobody who participated in governing France can plausibly describe the debt as if it were manufactured entirely by creditors against the wishes of elected governments.

The 1973 law did not suddenly ban free financing by the Banque de France

A recurring French story holds that a January 1973 law suddenly prohibited the state from borrowing from the Banque de France and forced France into the arms of private banks. The National Assembly’s own research on public debt says otherwise. Article 19 of the 1973 law still allowed advances and loans from the Banque de France under agreements between the finance minister and the governor, subject to parliamentary approval. [S8]

Budget reports provide the numbers. The ceiling for Banque de France support to the Treasury was 20.5 billion francs, and a tranche up to 10.5 billion francs could be non-interest-bearing. Those facilities therefore did not disappear in 1973. [S9]

More strikingly, the mechanism brought the amount of such support down to zero in 1982, and no new support was subsequently mobilised for the Treasury’s cash needs. That occurred under François Mitterrand, with Pierre Mauroy as prime minister and Jacques Delors at Finance. [S9]

The full legal prohibition arrived later, in the context of European monetary union. The 1993 reform prohibited direct overdrafts or credit to the Treasury and direct purchases of its debt by the Banque de France. Today Article 123 TFEU prohibits the ECB and national central banks from directly financing public authorities. [S8][S10]

One can argue that these rules should be changed. That is a legitimate political position. But the chronology matters: France had already built recurring deficits before the final legal prohibition, and the interest bill is not what creates the initial gap between spending and revenue.

The creditor did not write the French budget

Consider the simplest accounting identity. If a government spends 100 and raises 95, it must find 5. It can raise taxes, cut or defer spending, sell assets, create money where the legal and monetary system permits it, or borrow. France has repeatedly chosen borrowing for part of that gap.

Creditors therefore finance a difference created by public decisions. They can influence the price of financing, and high rates can worsen the debt dynamics. But they did not vote the French budget. This is the part of the story that disappears when debt is presented mainly as a plot by creditors.

Interest can of course become a snowball. Removing some interest costs may create fiscal space. But cancelling existing debt while keeping a structural deficit simply starts the accumulation again. Stock and flow are different problems.

Who actually holds French debt?

The phrase “filling the pockets of the privileged” also hides the structure of the creditor base. French public securities are held by banks, insurers, funds, central banks and other institutional investors, directly or indirectly representing households and savers in France and abroad.

The Banque de France reported that at the end of March 2026, non-residents held 55.9% of long-term public-sector debt securities, up from 54.6% at end-2025. [S14]

France therefore depends materially on international investor confidence. The Agence France Trésor plans €310 billion of net medium- and long-term issuance in 2026. At 31 July 2026, negotiable state debt stood at €2,881.9 billion with an average maturity of 8 years and 163 days. On 21 August the representative ten-year TEC 10 yield was around 4.08%. [S11][S12][S13]

That is a very different financing environment from the early 2020s, when France could issue some debt at exceptionally low or even negative yields.

A sovereign can default, but default is not costless

A sovereign state can legally choose not to honour some obligations. It cannot be repossessed like a house. But legal sovereignty is not economic immunity.

Mélenchon himself recognised the distinction in 2020. His parliamentary resolution proposed turning central-bank-held sovereign debt into perpetual zero-interest debt. In the explanatory text, a general payment default was described as a “disaster and chaos”. [S17]

IMF research by Luís Catão and Rui Mano found a persistent post-default premium in historical sovereign data. Their estimates should not be mechanically applied to modern France, but the mechanism is intuitive: once a borrower repudiates an obligation, future lenders demand compensation for the possibility that it could happen again. [S19]

For France, this matters because refinancing never stops. Bonds mature every year. The state still has to pay salaries, pensions, suppliers and social benefits. It also has to finance new deficits. Credibility is therefore not a moral lecture from markets; it is a parameter in the price and availability of future financing.

Mélenchon’s actual proposal should be described fairly

Critics should not pretend that Mélenchon is proposing to burn every French government bond held by private investors. His current argument targets, in particular, the share held inside the Eurosystem through the Banque de France. He has defended variants of this approach since at least 2020. [S1][S2][S17]

That is not the same thing as a generalized sovereign default. The accounting is also more complicated than a private creditor relationship because national central banks are public institutions and central-bank profits can flow back to governments.

But “nobody would notice” is still far too simple. A central bank that purchased government bonds created central-bank money in exchange and recorded the securities as assets. Cancelling the asset does not reverse the original money creation. It changes the central bank’s balance sheet, future income, loss-absorption arrangements and potentially the institutional relationship between fiscal and monetary authorities.

The legal barrier is also real under current law. Christine Lagarde wrote in April 2021 that cancelling public debt held by the Eurosystem would be incompatible with the treaties because it would violate the prohibition on monetary financing in Article 123. [S18][S10]

LFI can answer that the treaties should be changed. That is a coherent political choice. But then the proposal must be presented as a major redesign of European monetary governance, not a magic eraser.

The question after cancellation is the same: what happens to the deficit?

Suppose, for argument’s sake, that France succeeds tomorrow in cancelling or converting a large share of central-bank-held debt into perpetual zero-interest debt. The stock falls and the interest bill may fall. That could create genuine room for manoeuvre.

Then ask one question: if expenditure still exceeds revenue next year, what happens? France borrows again. If the deficit persists the following year, it borrows again. Over time, the cancelled stock is rebuilt.

That is why the distinction between debt stock and annual deficit flow is decisive. Cancelling part of the meter without repairing the leak can buy time; it does not repair the plumbing.

The Court of Auditors and OECD are warning about the same constraint

The French Court of Auditors says public finances remain among the most degraded in the euro area and warns that interest costs could exceed €100 billion by 2029. [S15]

The OECD’s June 2026 France survey reaches a similar conclusion. It describes putting the debt ratio on a sustainable path as a priority and notes that debt service rose from 1.3% of GDP in 2020 to 2.1% in 2025. [S16]

One may disagree with the policy prescriptions of these institutions. But dismissing the arithmetic does not make it disappear.

Protecting the welfare state also requires protecting fiscal capacity

This is the paradox in the rhetoric that debt exists to destroy social gains. Debt can absolutely protect social institutions during a crisis and finance their modernization. Yet once the interest bill becomes large enough, the debt itself crowds the room in which those institutions are financed.

A billion euros paid in interest cannot simultaneously pay nurses, teachers, police officers, judges, researchers, soldiers or infrastructure contractors. The point is not that interest is morally illegitimate. The point is opportunity cost.

A durable welfare state therefore needs durable financing. Fiscal discipline is not automatically anti-social. It can be the condition that prevents social promises from becoming impossible to honour later.

Fifty years of politically convenient postponement

France’s political system discovered a convenient mechanism: announce expenditure today, avoid raising all the taxes needed to finance it, avoid cutting another programme, and borrow the difference. Voters receive the benefit immediately while part of the bill moves into the future.

The left used that mechanism. The right used it. The centre used it. The Macron years used it, including for legitimate crisis responses and beyond. Citizens, too, have often rewarded promises while punishing the presentation of the bill.

Responsibility is therefore broader than a morality play opposing “the people” to “the privileged”.

What an international reader should take from the French case

France is a useful case study precisely because it is a rich, democratic state with deep capital markets, its own national central bank inside the Eurosystem and a very large public sector. Nothing in this article implies imminent insolvency. France continues to issue huge volumes of debt and investors continue to buy them.

The warning is about path dependence. A state can normalize deficits for decades because markets continue to refinance it. That can make the absence of a sudden crisis look like proof that there is no constraint. Then interest rates change, demographics change, defence needs rise, climate investment rises and the same stock of debt starts consuming budgetary room much faster.

The politically easy response is to blame the creditor or promise a one-off cancellation. The harder response is to decide which public services deserve more resources, which expenditures should be reduced or redesigned, which taxes can be justified, and what permanent balance the country is willing to defend.

What I expect from a presidential candidate

Jean-Luc Mélenchon is now a declared candidate for the 2027 French presidential election. He is entitled to campaign for a radical transformation of European monetary rules. [S22]

But a presidential proposal should answer the operational questions: exactly which securities would be cancelled or converted? At what legal level? What happens to Banque de France capital, income and loss-sharing? Which treaty provisions must change? How would euro-area partners react? How would the remaining deficit be financed? What rule would prevent the debt from being rebuilt after cancellation? What happens if investors demand a larger risk premium on the debt that remains market-financed?

Calling these questions “orthodox” does not answer them. They are the implementation details of the policy itself.

Debt is neither an invention of the rich nor a piece of meaningless paper

France currently has €3,536.1 billion of Maastricht public debt. The state plans €310 billion of net medium- and long-term issuance in 2026. More than half of long-term public-sector debt securities are held by non-residents. France’s representative ten-year yield is around 4%, and the Court of Auditors warns that annual interest expenditure could exceed €100 billion by 2029. [S4][S11][S13][S14][S15]

Against that background, presenting the problem mainly as the fault of creditors is, in my view, demagogic and historically selective. France asked investors to lend. Successive governments spent the proceeds. The public benefited, directly or indirectly, from the policies financed. The terms of financing, the ECB’s role and the treaties can all be debated. The original budgetary choice to spend more than was raised cannot simply be edited out of the story.

The most responsible message is therefore neither “repay everything immediately” nor “burn the bonds”. It is to admit that France has lived for half a century with spending above revenue and now has to decide democratically what it wants to finance, how it will finance it, and how much of today’s bill it is still willing to transfer to tomorrow.

Slogans can make debt disappear for a few seconds in a speech. Public accounting eventually brings the invoice back.

Primary sources and references

1. [S1] Jean-Luc Mélenchon, official Telegram channel, August 2026 posts on cancelling central-bank-held debt and on debt as a “deliberate strategy”. Open source ↗
2. [S2] TF1 Info, Vérif’, August 2026: context for the phrase “throw them in the fire” and clarification of the share of debt targeted by Jean-Luc Mélenchon. Open source ↗
3. [S3] L’insoumission, video interview with Éric Coquerel about his book Lâchez-nous la dette !, 21 October 2021. Open source ↗
4. [S4] Insee, Maastricht debt in Q1 2026: €3,536.1 billion, 117.5% of GDP, up €75.6 billion during the quarter. Open source ↗
5. [S5] French National Assembly, public-balance history since 1971: a 0.1% of GDP surplus in 1974 followed by annual deficits from 1975 in the reproduced series. Open source ↗
6. [S6] Insee, historical public-debt series: 20.7% of GDP in 1980, 22.0% in 1981, 31.1% in 1986, 33.3% in 1988, 46.0% in 1993 and 55.5% in 1995. Open source ↗
7. [S7] Insee, public debt and deficit 2012–2018: €1,892.5 billion and 90.6% of GDP in 2012; €2,258.7 billion and 98.4% in 2017. Open source ↗
8. [S8] French National Assembly, report on public-debt management and transparency: the 1973 law did not prohibit Banque de France advances to the state; the complete prohibition came in 1993. Open source ↗
9. [S9] French National Assembly, 2000 budget report: Banque de France–Treasury support ceiling of FRF 20.5 billion, including up to FRF 10.5 billion non-interest-bearing; support reduced to zero in 1982. Open source ↗
10. [S10] Treaty on the Functioning of the European Union, Article 123: prohibition on direct monetary financing of public authorities. Open source ↗
11. [S11] Agence France Trésor, indicative 2026 financing programme: €310 billion of net medium- and long-term issuance. Open source ↗
12. [S12] Agence France Trésor, negotiable state debt at 31 July 2026: €2,881.9 billion; average maturity 8 years and 163 days. Open source ↗
13. [S13] Agence France Trésor, key figures: TEC 10 at 4.08% on 21 August 2026. Open source ↗
14. [S14] Banque de France, issuance and holdings of French securities, Q1 2026: non-residents held 55.9% of long-term general-government debt securities. Open source ↗
15. [S15] French Court of Auditors, The State of Public Finances at the Beginning of 2026: interest costs close to €74 billion in 2026 and potentially above €100 billion in 2029. Open source ↗
16. [S16] OECD Economic Surveys: France 2026: public debt at 115.5% of GDP in 2025 and debt-service costs at 2.1% of GDP. Open source ↗
17. [S17] French National Assembly, resolution proposal no. 2914 (2020): conversion of debt held by the ECB into perpetual zero-interest debt; the text describes a payment default as “disaster and chaos”. Open source ↗
18. [S18] ECB, letter from Christine Lagarde dated 23 April 2021: cancellation of public debt held by the Eurosystem would be incompatible with the Treaties under Article 123. Open source ↗
19. [S19] IMF, Luís Catão and Rui Mano, Default Premium: historical study of sovereign post-default risk premia. Open source ↗
20. [S20] Olivier Blanchard, Public Debt and Low Interest Rates, American Economic Review, 2019. Open source ↗
21. [S21] IMF, Making Debt Work for Development and Macroeconomic Stability, 2022: debt is not inherently good or bad; outcomes depend on how and under what conditions it is used. Open source ↗
22. [S22] Le Monde, 3 May 2026: Jean-Luc Mélenchon formally announces his candidacy for the 2027 French presidential election. Open source ↗