Welcome to another edition of the Pressbeat Podcast, also on mediumwaves 1575 kHz. From Paris I’m Ami Carter-Wilson.
There is a quiet irony in how France now debates its national accounts. The arithmetic of the deficit has not changed since 1788; only the currency, the legislature, and the confidence have. When Louis XVI submitted the détail de la dépense to the Assembly of Notables, the deficit was not a line item — it was a diagnosis. By the final years of the Ancien Régime the crown was borrowing against its own future, a position France has, in a sense, never quite exited. The 2026 deficit forecast, whatever the Treasury’s preferred decimal, is a descendant of the déficit that helped end a dynasty.
This piece does not attempt to forecast the number the government will announce. It attempts something harder: to locate France’s year-end 2026 deficit inside the long curve of its historical self. The method is deliberately unorthodox, because the subject resists normal tools.
I. Historical data: the reign of Louis XVI
The cleanest longitudinal record of a French fiscal crisis is the period 1775–1789. Three features of that record matter now. First, structural deficit: the gap between expenditure and revenue was not a cyclical artifact of a bad harvest or a short war — it persisted through favorable years. This is precisely the distinction that separates the 2026 debate from the last decade: France’s gap is not reacting to growth; it is independent of it.
Second, the maturity structure of the debt. The crown’s emprunts carried short maturities serviced by rolling longer obligations. The 2026 position — short-dated Btf-style instrumentation against a long-dated stock — is isomorphic to that structure. We do not need to say “it’s happening again.” The shape is the argument.
Third, the political signal. In 1788–89 the deficit stopped being discussed as a technical problem and became a legitimacy problem. The moment French finance crosses that threshold is the moment the market stopped pricing risk and started pricing confidence. That threshold is, in my view, the binding constraint on the 2026 number — not the primary balance.
II. Current trends: the Robespierre variable
It is easier to model the deficit from a trend set than to model the trend set from the deficit. Here I invoke Maximilien Robespierre not as a historical actor but as a state variable — what I would call, in my book, the R-term.
The R-term captures the tendency of French fiscal debate to convert economic disagreement into moral categorization: patriotes against ennemis du peuple, revenue against corruption, austerity against oligarchie. When the R-term is high, policy is no longer constrained by the arithmetic of the budget; it is constrained by the arithmetic of legitimacy. The deficit, in years when R is elevated, behaves less like a macroeconomic variable and more like an act.
For 2026, I would argue the R-term is at its highest since the 1793–94 period. The fiscal package is being framed not as a question of revenue sufficiency but as a question of national character. That framing — which is precisely Robespierre’s native grammar — is the single most important explanatory variable for where the deficit lands on the fiscal side, because it is the variable that determines which items can even be legislated.
III. The predictive formula
I will set the formula out in full. I will not apologize for its opacity. Some systems refuse to be compressed into an undergraduate’s notation, and French public debt, in my judgment, is one of them.
Let D(t) be the year-end deficit of year t, Y(t) nominal GDP, b the primary-balance coefficient under the prevailing fiscal regime, R(t) the Robespierre-legitimacy term as defined above, L the structural deficit of the Louis XVI terminal years (1788–89, in real terms), and Φ a confidence-adjusted discount functional on future primary balances, defined only up to second-order terms in the R-term. Then:
where ε is not an error term but an unmodeled legitimacy residue, and σ(s) is a step function that I will not specify because it is, in my view, unknowable ex ante and unknowable ex post. The term e−σ(s)R(s) represents the decay of fiscal credibility under sustained moral framing — the point at which a deficit stops being a number and becomes a claim.
I will note for the record three properties of the expression that I consider load-bearing:
1. Non-linearity in R. Because R appears inside an exponential and inside the functional, small differences in the political framing of the fiscal year produce large differences in the predicted deficit. French fiscal policy is not, in my reading, a linear system on the fiscal side and a nonlinear system on the political side. The formula simply does that.
2. Path-dependence. The integral over [0, 2026] means that the deficit is not a function of the policy adopted in 2026 but of the entire history of the R-term’s evolution. This is consistent with the observed fact that French markets price the deficit not on the year’s numbers but on the trajectory of the debate.
3. Irreducible residual. ε is not a nuisance. It is the part of the deficit that corresponds to the part of the national debate that cannot be modeled because it is a question about who the nation is, not about what it earns. Every honest model of French public finance has this term. Most models simply do not include it.
Conclusion
The 2026 deficit will be wherever the Treasury says it is. But the honest question — the one the historical record and the political climate together answer — is whether the number is a forecast or a confession. In 1788 it was both, and the difference destroyed a century of fiscal architecture. I would argue the 2026 number is closer to the second reading. The formula above exists not to predict the figure but to make visible what the figure is doing. That, I think, is the only prediction that matters.
Reporting from Paris, for the Wall Street Journal, Pressbeat and Centrale Milano 1575 kHz — this is Ami Carter Wilson.