France's retirement age debate: the real problem isn't 67
France's pension watchdog projects a retirement age of 67.6 by 2070. Germany, the Netherlands, Italy, and Belgium are already there.
The real crisis is a labor market that pushes workers out long before they can collect.
Germany, the Netherlands, Italy, and Belgium already set their legal retirement age at 67. Denmark is at 68 and climbing toward 70 by 2040. In France, the mere projection of 67 and a half as a possible horizon in 2070 was enough to trigger an avalanche of outrage from unions, politicians, and the public. That reaction followed the early circulation of France’s 13th annual report from the Conseil d’orientation des retraites (COR) — the country’s independent pension advisory body, roughly equivalent to a joint congressional budget office focused exclusively on retirement finance — ahead of its official presentation on June 11, 2026.
The report contains a projection that encapsulates a demographic reality France has been reluctant to confront for decades.
This piece is not about whether French workers should work until 67. It is about why France is structurally unprepared to face that challenge while its neighbors have done so for a decade — and what the institutional data reveals that the political debate does not.
This image is used for illustrative purposes only.
At a Glance
The COR is not proposing a reform: the 67.6 figure is a purely arithmetical scenario, showing what balance would require if age were the only lever used. Confusing that calculation with a policy recommendation is the first flaw in the public debate.
For the first time since 1945, France’s natural population balance — births minus deaths — turned negative in 2025. A fertility rate of 1.56 children per woman, the lowest since 1917, has fundamentally altered the baseline projections. That is why the 2026 report differs from every previous one.
Fewer than 43% of French people aged 60 to 64 are currently employed. Raising the legal retirement age without transforming the labor market for older workers means extending a gap, not extending careers.
What the COR report actually says — and what the press simplified
The Conseil d’orientation des retraites (COR) is a consultative body attached to the Prime Minister’s office. It produces projections — not legislation. Its 13th annual report, presented on June 11, 2026, answers a specific technical brief: identify, across different scenarios, the level of each adjustment lever that would ensure long-term financial balance.
The figure of 67 years and six months corresponds to a purely arithmetical scenario — not a policy proposal — in which the only lever used is a gradual increase in the departure age, spread over 45 years: 64 years and 4 months in 2030, 65 years and 8 months in 2045, and 67 years and 6 months in 2070. This is not what the COR “recommends” — it is what the math requires if every other lever stays untouched. The COR does not set policy; it builds scenarios.
What makes this report different from every previous one is a fundamental shift in the underlying data. During its plenary session of April 16, 2026, the COR formally adopted the new demographic projections published by INSEE (France’s national statistics institute) in January 2026. These figures break sharply from previous assumptions: France’s fertility rate stands at 1.56 children per woman in 2025, compared to 1.8 in the previous reference scenario. It is the lowest level recorded since 1917. The direct consequence: for the first time since 1945, France’s natural demographic balance — births minus deaths — turned negative in 2025.
The COR had itself signaled this rupture in a letter published in February 2026, warning in unusually explicit terms that the financial diagnosis of the upcoming report “could be significantly revised compared to the June 2025 report.” For an institution that typically communicates in the cautious language of conditional projections, this was a clear alert.
The arithmetic of a pay-as-you-go system
France’s retirement system is funded on a pay-as-you-go basis: today’s workers’ contributions pay today’s retirees’ pensions. The sustainability of this model depends entirely on one ratio: how many working contributors per retiree.
In 2025, 17.2 million retirees receive pensions in France, at a ratio of approximately 1.7 contributors per retiree. That ratio, already below 2 since the late 1990s, is projected to fall further — to between 1.3 and 1.4 by 2070 — as the baby boom generation retires and low birth rates compound over decades. The total number of retirees could exceed 20 million by the mid-2040s.
This system already accounts for roughly €370 billion in annual expenditure, or 13.1% of GDP — one of the highest proportions among developed economies and well above the European average.
In March 2026, the COR released a separate macroeconomic study comparing four adjustment levers, each calibrated to achieve equivalent savings of €6 billion (approximately $6.6 billion at mid-2026 exchange rates). Three independent economic teams worked in parallel: the Treasury’s Mésange model, the OFCE (the French Economic Observatory) with its EmeRaude model, and the CEPREMAP research institute with CepreHANK. Their conclusions converged: raising contributions, cutting pensions, and restructuring financing are all classified as recessionary levers — they destroy jobs and suppress growth. Only raising the retirement age is classified as expansionary, with projections of 210,000 to 240,000 jobs created and a GDP gain of 0.7 to 0.9 percentage points.
The gap between legal age and labor market reality
This is where the structural contradiction becomes visible — one the COR’s macroeconomic report acknowledges directly. The simulation assumes that older workers will, in fact, remain in employment longer. But fewer than 43% of French people between the ages of 60 and 64 are currently employed.
The gap between the legal retirement age and the age at which French workers actually leave employment has been the blind spot of this debate for twenty years. A significant share of workers who “retire” after the legal age have spent one or more years in unemployment, disability benefits, or inactivity before gaining access to their pension rights. This gap — what some economists describe as a “precarity corridor” — could represent years of lost income and reduced contribution records for those caught in it.
Raising the legal retirement age without changing how employers manage older workers — who are often managed out of the workforce between 55 and 60 through negotiated departures — would likely extend this corridor rather than extend careers. The COR’s own long-term projections estimate that retirees’ standard of living, relative to the population average, could fall from 98.7% in 2021 to approximately 83% by 2070 at current legislative settings — a significant relative deterioration regardless of any change in retirement age.
France in a European context it watches from a distance
The European comparison is both illuminating and misleading. Illuminating because it reveals the scale of France’s structural lag. Misleading because retirement systems are not directly comparable across countries.
In 2025, Germany, Belgium, the Netherlands, and Italy have set their legal retirement age at 67. Denmark is at 68, with a legislated path to 70 by 2040. These countries integrated years ago the principle of progressive, automatic age adjustments tied to life expectancy — a mechanism that removes the decision from the political calendar.
Sweden has developed the most sophisticated model: the departure age adjusts automatically based on life expectancy and system performance. This is not a political choice revisited each election cycle — it is a parametric rule embedded in the system’s architecture.
France currently sits at 62 years and 9 months as its minimum departure age, with a trajectory toward 64 that has been partially suspended through 2028 under the social security financing act for 2026. The gap with Germany and the Netherlands is approximately 4 to 5 years on the legal retirement age, and around 3 years on average effective labor market exit ages.
This gap could reflect a political tradition in which retirement is treated as a universal social right earned through collective solidarity — a framework that would make every parametric adjustment far more politically costly in France than in the Nordic or Germanic countries. Sweden’s indexed system works in part because it was negotiated in the 1990s during a severe fiscal crisis, with cross-party consensus that has not been replicated elsewhere.
Analysis
The long trajectory: a slope, not a cliff
The current French debate creates the impression of a radical break. It masks a longer trend: since the Balladur pension reform of 1993, which extended required contribution periods, through the Woerth reform of 2010, the Touraine reform of 2014, and the 2023 reform that raised the legal retirement age from 62 to 64, each decade has brought its adjustment. A legal retirement age of 67 in 2070 is not a discontinuity — it is the mathematical continuation of a trajectory already underway for thirty years.
The mechanics of power: who decides, and when?
The COR is a tripartite body — unions, employer federations, government representatives, and parliamentarians — whose members do not necessarily agree on the report itself. The plenary vote of June 11, 2026 is unlikely to result in unanimous adoption: several union organizations have already signaled their opposition and announced they will not endorse a report pointing toward a higher retirement age. This matters institutionally: a report rejected by its own members carries far less weight in the political debate than a consensus document, and gives each political camp license to dismiss or weaponize its findings selectively.
Eighteen months before the 2027 presidential election, the stakes are high. Its projections will inevitably become electoral terrain, with each camp appropriating figures that suit their platform — without necessarily maintaining the distinction between a technical scenario and a policy choice.
The fundamental question: can the legal age be raised without transforming the labor market?
This is the unresolved tension the macroeconomic models raise without answering. If fewer than half of 60-to-64-year-olds are employed today, and if companies continue managing their older workforce out through negotiated early exits at 57 or 58, raising the legal retirement age to 65 or 67 does not mechanically create jobs for older workers. It creates additional years of unemployment insurance or disability benefits before pension rights open.
The COR’s models produce projections of retirement expenditure. They do not model the shift in costs toward unemployment insurance or the healthcare system that a higher retirement age would likely produce without a complementary employment policy for older workers. That transfer of costs — from one public account to another — is the gap the debate is not having.
Can a pay-as-you-go retirement system designed for a ratio of four contributors per retiree survive a ratio of 1.3 — and if so, at what cost, and paid by whom?
The bottom line
France is debating 67 in 2070 as though it were a distant fiction. But the demographic rupture documented by INSEE in January 2026 — negative natural balance, fertility at its lowest in a century — is not a projection. It is the present.
The deeper question the COR’s report raises without explicitly asking is about the model itself. Before 2027, it is unlikely that presidential candidates will answer that question in full. But it will be there, in every debate, waiting.
Sources: Conseil d’orientation des retraites — 13th annual report (plenary session June 11, 2026) · COR — macroeconomic study on pension adjustment levers (March 2026) · INSEE — demographic assessment January 2026 · Vie Publique — synthesis on birth rate and pensions (February 2026) · Eurostat — EU demographic projections · France Info / France 2 evening news (June 9, 2026)