France produces more, but passes on less. In 2023, once 201 billion euros in climate damages and carbon budget consumption were deducted, French net savings fell to minus 133 billion euros, according to new accounts published by INSEE on November 5, 2024. This figure does not say the economy is collapsing. It says that standard indicators do not cover certain environmental and social dimensions that expanded accounts aim to fill.
The essentials
- In 2023, French net savings become negative (-133 billion euros) once climate damages and carbon budget consumption are included, according to INSEE.
- The adjustment accounts for 201 billion euros: standard accounts do not make climate costs visible or isolable, and do not track the decline in climate assets passed to future generations.
- Steering with GDP alone amounts to ignoring a real liability: wealth produced today is partly financed by a draw on tomorrow’s collective assets.
- The institutional question is open: integrating these expanded accounts into budget decisions requires choosing between several measurement architectures with divergent political implications.
- The European Commission and several member states are moving toward common frameworks for expanded accounting, but the timeline and constraints remain to be defined.
INSEE’s measurement method
Net savings, in ordinary national accounts, measures what remains once physical capital maintenance is deducted. It is already a more demanding indicator than gross GDP: it removes the depreciation of machines, buildings, and infrastructure. In 2023, French net savings were positive, at a few tens of billions of euros.
INSEE’s expanded accounts take an additional step. They introduce two additional asset classes: climate capital, valued at the social cost of carbon, and the carbon budget, valued by the Climate Action Value (VAC), which shrinks with emissions compatible with the French trajectory. Together, these two adjustments represent 201 billion euros in 2023. Result: net savings shift into negative territory, at minus 133 billion.
The method is transparent and documented. INSEE makes visible a liability that international accounting conventions do not yet systematically capture. Valuing carbon at social cost—an estimate of what each ton emitted will cost future generations in damages—remains a methodological choice open to discussion. With INSEE’s retained valuations, the adjustment is negative and substantial, without proof that it would remain so regardless of the carbon value.
This work is part of a broader international program. The OECD, the World Bank, and the UN System of Environmental and Economic Accounts have been working for years on similar frameworks. A limited number of countries publish these adjustments within the institutional framework of their national accounts, which gives this result a different weight than an academic study would have.
French data through the lens of Diane Coyle’s compass
British economist Diane Coyle has long defended a thesis simple in its implications and difficult in its consequences: you cannot steer what you do not measure. In The Measure of Progress, she shows that our statistical systems were designed for an industrial economy of the mid-twentieth century. They remain blind to digital assets, non-market services, and above all to natural capital. GDP records plastic production and pollution cleanup expenditures with the same satisfaction. It sees neither as a loss of assets.
INSEE’s expanded accounts provide a precise French translation of this criticism. The figure of minus 133 billion euros is a political figure: it indicates, according to this experimental method, that standard accounts do not make visible certain implicit climate costs affecting future assets.
A competing reading exists, advanced notably by liberal economists concerned with regulatory burdens on competitiveness—think of voices like Philippe Aghion or Jean Tirole on the design of climate instruments. It does not contest the reality of climate damages. It contests the way to account for them: valuing carbon at social cost implies a discount rate, that is, a judgment on what the future is worth relative to the present. This rate is as much a philosophical choice as a technical one. A high rate reduces future damages and lowers the adjustment; a low rate amplifies them.
The figure of 201 billion depends on this parameter.
This tension is real and makes the approach more exacting. Publishing the valuation assumptions retained and their alternatives, which INSEE does in its methodological documentation, enables debate. A contestable and transparent indicator is preferable to an indisputable one because it is invisible.
The significance of -133 billion for public budgets
The practical scope of this figure depends on what you decide to do with it. Presented alone, it says nothing about the policy to pursue. It says something about policies already pursued: in standard national accounts, climate costs induced by emissions are not made visible or isolable.
Take a concrete example. A decision to invest in an additional highway generates measurable GDP growth, recorded employment, and real tax revenues. It also generates emissions that count against the national carbon budget, and thus against expanded savings. In ordinary accounts, only the first effect is visible. In expanded accounts, both are.
The decision-making trade-off is not mechanically different; other criteria come into play, but it is better informed.
The same logic applies to prevention spending. Renovating a building to make it less energy-intensive costs public funds, reduces the ordinary fiscal deficit, but improves expanded savings. In classical accounts, only the cost is visible. INSEE publishes more internationally on these topics but still has little weight on the accounting standards that structure European decision-making.
The governance problem is there: European budget rules, the Stability Pact, its deficit and debt criteria, are built on classical national accounts. A resilience investment can worsen the overall public finance balance while increasing recorded produced assets on the balance sheet; climate capital nonetheless remains outside the standard framework. The structural incentive points in the wrong direction.
Three obstacles to political integration of these accounts
Naming the specific obstacles is better than vague appeals to complexity.
The first point of debate concerns valuation. INSEE retains 172 €/tCO2e in 2023, with a broad range of methodological uncertainty. France’s High Carbon Authority retains different values from those of the European Union, which in turn differ from those used by the World Bank. These gaps can shift total adjustment by several tens of billions. Without an international convention on this valuation, comparing expanded savings between countries remains difficult.
The second obstacle is institutional. The indicators that steer budget decisions change only through treaty or regulation. The 2024 reform of the European budget framework introduced multi-year expenditure trajectories, but without integrating climate accounting. Bringing expanded accounts into the Maastricht criteria or Commission rules would require lengthy negotiation, with very different coalitions of interests among member states, the still heavily carbonized economies of Eastern Europe naturally resisting rules that would mechanically worsen their expanded balance.
The third obstacle is democratic in the technical sense. An indicator that enters political decision-making must be understandable by those who vote. Expanded savings is a concept economists handle, not yet a public reference point. The analogy to public debt is useful: it took decades for “deficit” and “debt” to become tangible political realities for voters. Expanded accounts are in their infancy.
The indicators that institutions will steer by from 2030 to 2040
INSEE’s accounts have both descriptive and prospective scope. If these indicators enter decision-making frameworks by 2030-2040, the trade-offs regarding investments in natural capital and short-term spending will be the first to change.
The most direct concerns the evaluation of public investment policies. France launched the France 2030 plan, which directs tens of billions toward industrial and technological sectors. These investments are evaluated by their returns on GDP and employment. If the evaluation also integrates their impact on expanded savings, their carbon footprint, their contribution to or draw on the national climate budget, the ranking of priority projects could change significantly. Investments with modest GDP returns but zero or negative carbon footprint would rise in the ranking.
The second potential change concerns budget rules. The European Commission is working, within its green accounting agenda, to integrate sustainability indicators into economic surveillance of member states. The 2026 annual report on employment and social development in Europe, the ESDE, explores how expanded accounts could inform national recommendations. If a member state presents persistently negative expanded savings, this could justify specific recommendations, distinct from those related to ordinary deficit. The mechanism remains to be defined, but the movement has begun.
The most instructive signal to watch is in sovereign bond markets. Rating agencies have begun integrating climate criteria into their sovereign assessments. Persistently negative expanded savings could weigh on perceptions of a state’s long-term sustainability, transforming the incentive to adopt these accounts from a political argument to a financial one. France is not yet in this situation, its rating remaining solid, but this vector of diffusion is real for the coming decade.
The most tangible progress path in the years ahead is probably international standardization. The OECD and the United Nations regularly publish revisions to the environmental accounting framework. Each revision that becomes an international statistical norm increases pressure on states that have not yet applied it. France, by publishing these accounts within its official national accounts, takes a forward-looking position that can weigh on this standardization, much as it did, with others, on corporate carbon accounting through European regulation.
South Africa is experimenting with similar approaches to audit its economy without surrendering data sovereignty. Economies most dependent on natural capital have a direct interest in seeing it appear in national accounts. Convergence between advanced economies and some emerging ones on this question indicates that standardization is progressing at a steady pace.
The problems the next generation of indicators must solve
INSEE’s minus 133 billion is a first result, not an endpoint. Several methodological questions remain open and condition the political credibility of these indicators.
The main one is symmetry. Expanded accounts charge the consumption of climate capital and carbon budget, treated as additional assets, but do not yet integrate the full range of natural capital as an asset. A sustainably managed forest, a preserved coastline, a reconstituted aquifer: these assets do not yet appear in French accounts with the same rigor as emissions on the liability side. Complete accounting should record both sides of the balance sheet. INSEE indicates this in its documentation: work is underway.
The second question is frequency. Publication of expanded savings covering 2023 came on November 5, 2024. A three-year lag makes the indicator difficult to use for annual budget debates. Accelerating the production of these accounts, as has been done for energy accounts or tourism satellite accounts, is a technical condition for their political usefulness.
The third question is international comparability. Geographic capacity to produce expanded accounts is very unequally distributed among states. Without harmonization of methods, each country can produce its own figures with its own assumptions, making any comparison misleading. This is one of the strongest arguments for rapid standardization by international organizations.
France has produced a result. It now has to decide whether to make it a governance tool or an academic exercise. The answer depends less on economists than on the institutions that decide which indicators they will impose on themselves to examine.
