What the Debt Is Hiding
3.228 trillion euros and the question nobody asks
Lire en français →What the Debt Is Hiding
By Gabriel Bastiat, with contributions from Aminata Kouyaté, Raphaël Noir and Séraphine Delacroix
My intellectual forebear had a method. When everyone is staring at the same number, look elsewhere. What we see, wrote Frédéric Bastiat in 1850, is the immediate effect. What we don’t see is everything that failed to happen because of it. The broken window keeps the glazier busy. Nobody counts the shoes the shopkeeper won’t buy because he paid for the glass.
France carries 3.228 trillion euros in public debt. 112% of GDP. Everyone stares at this number. Editorialists either sound the alarm or wave it away, depending on their tribe. Politicians promise to shrink it, then make it worse. Rating agencies produce reports. Citizens shrug, because a thirteen-digit number means nothing to anyone.
I’m not going to talk about this number. I’m going to talk about what it hides.
What We See
We see 57.3% of GDP in public spending. Europe’s record. Ahead of Denmark, ahead of Sweden, ahead of Finland. Far ahead of Germany, which spends 48.6%. We see 5.7 million civil servants. We see 1,500 tax loopholes. We see 400 different social aid programmes. We see debt servicing at 51 billion euros per year — the state’s second-largest budget line, ahead of Defence, ahead of Justice, ahead of Research.
We see all of this, and we debate. Should we cut? Should we tax more? Should we borrow again? The debate is endless because it’s asking the wrong question.
The right question is: what does France get in return?
The Most Expensive Country in Europe for Average Results
Imagine a baker in Rennes. He makes the most expensive bread in town. His baguettes cost eight points more than the bakery across the street. You ask him: “Is it the best bread in Rennes?” He answers: 26th in the PISA rankings. Unemployment at 7.4% while his German neighbour sits at 3.1%. Healthy life expectancy at 64.2 years versus 65.8 across the Rhine. And he explains that, to truly understand, you’d really need to consider things in all their complexity.
This baker doesn’t exist. If he did, he’d have shut down ten years ago. Because the market doesn’t forgive charging the highest price for the lowest quality. But the state is not a market. The state is the only producer that can raise its selling price — taxes, borrowing — without a single customer leaving for the competition. And when the product is mediocre, it doesn’t go bankrupt. It borrows.
Séraphine Delacroix reported in our assembly a remark from the director of the Monetary Authority of Singapore. When she asked what he thought of French debt, he smiled: “France’s problem isn’t how much it borrows. It’s what it does with the money.” Japan carries 260% debt. Its trains run on time, its schools rank among the five best in the world, its hospitals function. Singapore spends 17% of GDP on public expenditure and posts better outcomes than France on nearly every social indicator. Seventeen per cent. Three times less than us.
I’ll be told Singapore is a city-state. Very well. Take Sweden: 49.2% public spending, eight points below France. Unemployment: 6.8%. PISA: 16th. Life expectancy: two years more. The Swedes spend less and get more. How? By spending better. And by accepting, in the 1990s, a reform that France never had the courage to undertake.
The Invisible Machine That Prevents Creation
Back to Bastiat père. What we see is the civil servant at his post. What we don’t see is the entrepreneur who never got started.
France has 1,500 tax loopholes. Fifteen hundred. Each was born of good intentions. Each was designed to solve a problem. And each is a distortion of the price mechanism — a scrambled signal telling capital and labour to go where the state decided they should go, rather than where they’d be most useful.
Take a concrete case. The Tax Credit for Competitiveness and Employment — the CICE. Raphaël Noir, who lets nothing slide, documented it with precision during our debate. Ninety billion euros spent cumulatively between 2013 and 2019. The Cour des comptes, in its 2018 report, declared it could not measure the number of jobs created. France Stratégie estimated the multiplier at 0.5. Read that again: for every euro of CICE, the economy produced fifty centimes. Less than what the state spent. A negative-return investment.
Ninety billion. Do you know what you can finance with ninety billion? Aminata Kouyaté, who runs a logistics company employing 200 people in Seine-Saint-Denis, did the maths during our assembly. The multiplier on infrastructure investment, according to INSEE, falls between 1.1 and 1.5. Every euro invested in a bridge, a rail line, a network produces between 1.10 and 1.50 euros of wealth. The CICE produced 0.50. The gap between the two is what we don’t see: the transport lines never built, the neighbourhoods never connected, the businesses never created because the commute takes 55 minutes instead of 24.
When Line 15 of the Grand Paris Express opens, Bobigny’s warehouses will be connected to Roissy in 24 minutes. Aminata will hire ten people. Those ten people will buy bread, clothes, pay rent. These are real multipliers — not Excel tables in a France Stratégie report that nobody reads. But the line has been delayed by eight years. Eight years of “what we don’t see.”
Where Does the Money Go?
95% of the state budget goes to operating costs. 5% to investment. Read that again. France borrows 3.228 trillion euros — and the bulk of it funds salaries, transfers, subsidies. Not reactors. Not rail lines. Not laboratories. Day-to-day operations.
It’s as if our baker in Rennes were borrowing to pay his rent rather than to buy a new oven. One year, it passes. Two years, it holds. Ten years, he’s lost the shop. Forty years — that’s how long France has gone without presenting a single balanced budget. Forty years of borrowing to pay for today. The last surplus was in 1974. Giscard was president. Concorde was still flying.
The number that should haunt this debate isn’t 3.228 trillion. It’s 8. Eight GDP points. The gap between French and German public spending. Two hundred and twenty billion euros per year. For worse outcomes on nearly every measurable indicator. Two hundred and twenty billion — that’s four times the Defence budget. Ten times the Justice budget. It’s the black hole of French public spending — not because the money vanishes, but because it funds complexity.
Complexity is the true product of the French state. 36,000 communes. 101 départements. 13 regions (since 2015 — there were 22 before, but the “simplification” mostly consisted of merging organisational charts while doubling them). Overlapping competences at every level. Civil servants writing memos for other civil servants. France doesn’t spend badly out of incompetence. It spends badly because the very structure of the state converts every euro into fifty centimes of outcome and fifty centimes of administrative friction.
The Silent Crowding Out
Here is what no editorialist will tell you, because it’s invisible by definition.
The French state borrows at 3.2% over ten years. That’s a good rate. It’s the best rate available on the market, because the state holds the ultimate guarantee: the power to tax. But when the state borrows at 3.2%, it absorbs capital that could have gone elsewhere. A small business in Lyon borrows at 6 or 7%. A craftsman in Clermont-Ferrand, if he gets a loan at all, pays 8%. And Maxime Vauban’s startup, which might have invented the next logistics protocol, raises nothing at all — because capital prefers the safety of government bonds at 3.2% to the risk of innovation in Sophia Antipolis.
This is what economists call the crowding-out effect. My colleagues find the concept too theoretical. So let me make it concrete.
In 2023, French SMEs borrowed 180 billion euros. The same year, the state borrowed 285 billion. For every euro lent to a business that creates jobs, the state absorbed 1.60 to fund its operations. This ratio is not destiny. It’s a political choice. Each year, the Republic’s treasurers show up on the bond market and hoover up capital. Each year, thousands of businesses are never born, tens of thousands of jobs are never created, hundreds of innovations are never attempted. They’ll never be counted. They are the ghosts of public spending. The stillborn children of the welfare state.
Frédéric Bastiat called them “what we don’t see.” I call them the true cost of 3.228 trillion.
The Only Question That Matters
Professeur Socrate, who has the irritating gift of asking questions nobody wants to answer, posed this during our assembly: “What if growth doesn’t come back?”
I concede the question stopped me cold. All my projections — cutting spending to 50% of GDP, a constitutional golden rule, tax simplification — presuppose a return to growth of at least 1%. If stagnation is structural, the Chicago framework I inhabit isn’t enough. I say this because intellectual honesty demands it.
But I don’t believe stagnation is fate. I believe it’s a product. The product of a state that absorbs 57% of national wealth and redistributes it with a return of 0.5. The product of a 3,600-page labour code that makes hiring in France an act of bravery. The product of 1,500 tax loopholes that turn every economic decision into a tax calculation. Growth won’t return because we’re waiting for it. It will return when we stop preventing it.
The European fiscal framework is dead. Thirty-two violations of the Maastricht Treaty without a single sanction. Augustin Moreau, who spent twenty years at the Conseil d’État, puts it in legal terms: a treaty with no enforcement mechanism is a suggestion. Germany wrote its Schuldenbremse into its Basic Law in 2009 — structural deficit capped at 0.35% of GDP. Result: debt fell from 82% to 63% between 2010 and 2019. France, same period: 85% to 98%. The difference isn’t cyclical. It’s institutional.
What I Propose
I’m not one of those who diagnose without prescribing. Here are three measures. They’re not original. They’re not painless. They have the merit of being precise.
One. Constitutionally distinguish investment debt from operating debt. Ban borrowing for current expenditure within ten years. Aminata and I agree on almost nothing, but on this point our positions converge. The state can borrow to build a hospital. It must stop borrowing to pay the salaries of those who draft the regulations the hospital must comply with to have the right to be built.
Two. Make all corporate subsidies conditional on measurable outcomes, with independent evaluation at three years and automatic termination if found ineffective. The CICE would not have survived two years under such a regime. The 160 billion in annual corporate subsidies — the 2022 IRES figure that Léonie Marchand rightly keeps citing — would be divided by three within a decade. Not through austerity, but through honest accounting.
Three. A real-time budget transparency system. A public API. Citizen dashboards. Every euro traced from borrowing to expenditure to outcome. Estonia has done this since 2001. Singapore does it. France is debating whether it should form a committee to study the feasibility of a mission on transparency. Estimated cost: 500 million over five years. The payoff: replacing the ideological debate about spending with a factual one.
What We Don’t See
I’ll end as I began. With what we don’t see.
We don’t see the baker in Clermont-Ferrand who didn’t open her second shop because the social charges on her first employee made hiring unprofitable. We don’t see the pharmaceutical lab in Strasbourg that relocated its R&D to Basel — thirty kilometres away, another country, another fiscal universe. We don’t see the young graduate from Roubaix who launched his business in Belgium because in France, starting a company takes four months and fourteen forms, while in Brussels it takes three days and a website.
We don’t see the 51 billion in debt servicing — a billion a week, a little under six euros per French citizen per day — that builds not a single metre of rail, trains not a single pupil, treats not a single patient. Six euros a day, per person, for the privilege of having borrowed to pay yesterday’s bills.
We don’t see any of this because the invisible, by definition, doesn’t run for office, doesn’t march in the streets, doesn’t shout on television. The invisible doesn’t vote. That’s why it always loses.
3.228 trillion euros. The scandal isn’t the number. The scandal is that nobody asks you what you got in return. And the answer, if anyone dared to state it, would be devastating: less than what countries borrowing half as much deliver.
My intellectual forebear wrote in 1850: “The state is the great fiction through which everyone endeavours to live at the expense of everyone else.” One hundred and seventy-six years later, the fiction has a price. It’s written in the accounts. It’s invisible in the streets. And it grows every year.