Official figures and this site's calculations
On this site
This site is an experiment. It brings together official figures on where French public money comes from and goes, who pays it and who receives it, who owns the country's wealth, and what is known of the effect of public spending, and it adds calculations and evaluations of its own. Below, each finding that rests on them is marked “Calculated by this site”. Behind every number, a panel shows the source it was read from or the calculation that produced it. Reports of errors and other feedback are welcome at github.com/samoht/budget/issues.
Estimated: a model, an imputation or a surveyHypothetical: a scenario or a forecastUnmarked figures are published as observed.
Where does public money come from and go, and who pays and who receives it?
Revenue and spending by purpose; what each group pays and receives, by living standard and by age; France beside Germany, the UK and the US.
Who carries the tax?
What taxes fall on, the share of wages in value added, and the tax rates of the richest households.
How much public money goes to companies, and what is known of its effect?
Public support to companies under each official count, the largest schemes and groups, and how long tax expenditures last.
Who owns what?
What France owns and owes, who holds it, what it earns and passes on, and the taxes proposed on the top, beside their costings.
Does it work?
For each line of public money, whether its effect has been measured, and this site's own evaluations.
Method and sources
Where the figures come from and how they were computed, with the files to download; who makes the site, and the corrections made to it.
What each tenth paid and received in 2023
In 2023, the latest year of Insee's distributional accounts, recalibrated by this site to Eurostat's totals for general government with a layer for what goes to and comes from outside France, counted per person and per month, the poorest tenth received €1,730, estimated in pensions, benefits and public services, and paid €380, estimated in taxes and contributions. The richest tenth received €2,150, estimated and paid €4,930, estimated. The difference is pensions, which follow past contributions and are counted here as transfers: without them the richest tenth received less than the poorest. Public services are counted at what they cost to provide.
Calculated by this site. Who pays and who receives
Top tax rates in 2016
Counting the corporate tax of the companies households control, the Institut des politiques publiques finds a tax rate of 46%, estimated at the entry to the top 0.1% and 26%, estimated for the 75 richest households, in 2016, before the 2018 tax reforms. Insee does not publish what the 27,000 people with the highest living standard paid in 2023: this site's extension of Insee's accounts along the top of their ranking, with its range and its assumptions, is under Who pays and who receives.
What people paid and received by age, 2023
With every levy and transfer of 2023 split by each person's own age, people aged 40 to 49 paid €1,290, estimated a month more than they received, and people aged 80 and over received €2,940, estimated a month more than they paid. The split follows published age profiles. It is one year's snapshot by age, not a lifetime account: the same people paid in when they worked, and pensions are the counterpart of those past contributions. Counting pensions as deferred labour income instead of transfers, Insee finds that the richest tenth's income before transfers falls from 26, estimated to 14, estimated times the poorest tenth's.
Calculated by this site. Who pays and who receives, by age
How wealth is spread, 2024
In 2024, the richest 10% of adults held 59.9%, estimated of net wealth, and the poorest half 4.8%, estimated. At the top of the distribution, 530 adults held 5.6%, estimated, €1.5bn, estimated each on average. About €300bn, estimated is passed on each year by inheritance and gifts; the duties on it raise €20.8bn, 6.9%, estimated.
Where taxes fell, 1995 and 2024
of taxes and contributions came from income from work in 2024, and in 1995. Employees' share of value added was in 1995 and in 2025.
Who lent to general government, 2023 to 2025
Of each year's net new public borrowing, lenders outside France took up 104.2%, estimated in 2023, 73.9%, estimated in 2024, 91.8%, estimated in 2025. The Banque de France's share was negative each year, as it let bonds mature, which general government repaid by borrowing from others. The Banque de France paid the State nothing for 2023 and 2024, having made losses. The Government forecasts public debt at 119.3%, forecast of GDP at the end of 2026, and interest on it at €79.2bn, forecast.
Calculated by this site. Who lends to general government
Public support to companies under each official count, 2023
Official counts for 2023 run from €82bn, estimated to €211bn, estimated. France has 465 tax expenditures, costing €89.4bn; 89.1% of that cost is in measures with no end date. According to the Cour des comptes, of the 35 evaluations the government scheduled for 2020–2023, 4 were carried out.
What this site's evaluations measured, 2008 to 2026
Two changes are descriptive. In January and February 2026, while the apprenticeship hiring aid for higher-education contracts had lapsed until the year's finance law and decree, those contracts changed by −38.1%, estimated against secondary ones; since the 2023 pension reform, employment at 60-64 has risen by 5.5 pts, estimated more than at 55-59. Neither design can attribute them to the aid or the reform alone, since large firms had no aid at any level in those months, employers could wait for its return, since July 2025 employers have paid a flat €750 towards the training of each contract at niveau 6 or above, and the gap between the two age bands also moved in 2005-2010, with no change in the legal minimum age. A result counts here when its 95% interval excludes zero and it passes its design's checks: no gap before the change, and a placebo that finds nothing. Two do, both in the State's grants to communes. At the cutoff of the fraction cible of the rural solidarity grant, €324m, estimated in 2019, where the grant comes and goes from year to year, communes saved the first-year euro in 2016-2019: self-financing rose by 1.36, estimated euros per euro, by regression discontinuity. A commune must balance its operating budget and may borrow only to invest, and a receipt that may not return the next year is one it can set aside. At the DETR cap, equipment spending rose by €50, estimated per inhabitant in the first year against each commune's own past, where the placebo finds +€0.5, estimated; measured in level the jump is €54, estimated, half of which appears where no rule changes, and over three years the design cannot tell the State's leverage from zero. Other results exclude zero but fail a check or have none. In the rural tax zone, private jobs changed by −7.4%, estimated and employers fell too, of the order of the pairs' drift before 2016; population, migration and the labour force fell as their unclassified neighbours' did, and income per household fell in the placebo before 2017 too. Industrial jobs in the industrial territories changed by −6%, estimated, a census difference that no placebo can check. Action cœur de ville's towns lost employers, shops and services, but no balanced match was found for them, and the services around the small towns of Petites villes de demain were falling before the programme. Where Pinel closed in zone B2 in 2018, flats authorised and the median price of all dwellings sold fell against zone B1, but the estimators find gaps of the same size where no rule changed, and zone A shows a price fall against B1 too. Communes under the DETR cap borrowed more in level, but not against their own past. France Relance winners published accounts less often than their matches, in a design whose placebo fails in 2018. The communes added to the vacancy tax in 2024 saw vacancy fall against never-taxed communes no more than the communes whose status did not change, and their recorded vacant dwellings rose as the files they are counted from changed. Against the schemes' own aims, the other outcomes measured move by amounts these estimates cannot tell from zero: researchers with the research tax credit, 9.4%, estimated above a synthetic France in 2008-2012, a result that cannot be told from zero (p 0.10, the smallest ten countries allow); business research; firms' turnover and margin; pupils' results; jobs in the small towns and the industrial territories; the vacancy rate in the communes added to the vacancy tax, against the communes taxed since 2013; and the flats authorised where the zoning moved communes up.
Calculated by this site. Evaluated by this site
Which public money has a causal study, 2024
Single measures with their own cost and a causal study, 13 of them, hold 6.1% of public money in 2024, €1,845bn; €77.6bn of that cost is the general cuts in employer contributions, counted in 2024 for all compulsory schemes, whose studies do not cover the current scheme, and without them the share is 1.9%. Of the 155 official goals set for its lines, a causal study measures 50 and touches 11 more only indirectly. Counting a broad line in full as soon as one study touches one of its rules gives an upper bound, 82.3%: defence counts because of studies of the end of conscription. Three large lines show measured effects, in both directions. The general cuts in employer contributions on low wages (€77.6bn) had a significant effect on employment when they were introduced in the 1990s, France Stratégie's Bozio-Wasmer mission finds, although the IPP detects none on employment from the conversion of the CICE into a cut in employer contributions in 2019. Halving classes in priority-education primary schools, a small part of the €42.2bn spent on pre-primary and primary education, pays for itself in the long run if its early gains persist, on a calibrated projection (Fajeau, Grenet, Laveissière and Leonetti, 2025) that the Conseil d'analyse économique draws on; the effects measured in France are modest and partly fade. Of reduced first-grade classes, Bressoux, Lima and Monseur (2019) write « At the end of Grade 2, the effects have faded »; of the halving itself, Merle (2022) finds « effets modérés, voire nuls ». Housing benefits, in a line of €21.1bn, push private rents up: when they were extended to new households in the early 1990s, between 50% and 80% of the benefits low-income households received went into higher rents, Fack (2005) estimates in Insee's Économie et statistique. Each line's studies and their findings are under Impact, leakage, permanence.
Calculated by this site. What is known of its effect
Public spending in France and three other countries, 2024
In 2024, public spending was 57% of GDP in France, against 49.4% in Germany, a difference of 7.6 pts. Old age, mainly pensions, accounts for 3.5 pts of it: 13.4% of GDP in France against 9.9% in Germany; social protection as a whole for 3.3 pts. On the OECD's measure of net total social spending, which counts private social benefits too and nets out the taxes levied on benefits, France spent 31.9%, estimated of GDP in 2021 and the United States 33.2%, estimated; their gross public social spending was 32.7% and 21.6%. In 2019, before the pandemic, their net total social spending was 30.1%, estimated and 29.2%, estimated. The private part includes the health plans employers finance, which the OECD notes are important in the United States.
France, 2025 to 2027
2025, provisional
GDP
on 2024, nominal
Public revenue
of GDP
Public spending, tax credits included
of GDP
Deficit
of GDP
Public debt
of GDP
Public means general government: the State, local authorities and social security, consolidated. 2025 figures are Eurostat's, provisional.
2026, as forecast by the Government
GDP
on 2025, nominal; in volume
Compulsory levies
of GDP, against in 2025
Public spending, tax credits excluded
of GDP, against and in 2025
Deficit
of GDP
Public debt
of GDP
2027, as forecast by the Government
GDP growth
in volume; nominal
Compulsory levies
of GDP, against in 2026
Public spending, tax credits excluded
of GDP, against in 2026
Deficit
of GDP, against in 2026
Public debt
of GDP at the end of 2027, against a year earlier
Forecasts of the Government's draft budget for 2027 (projet de loi de finances), September 2026, as reproduced in the opinion of the Haut Conseil des finances publiques of 25 September 2026. The Government counts spending without tax credits and revenue as taxes and social contributions only, so these two rows are not on the basis of the 2025 row above, and the difference between them and it is not a change; the 2025 figures beside its spending and levies are the Government's own, on its basis. The Haut Conseil judges the forecast deficit for 2026 plausible, though significant uncertainties remain on spending and revenue, and the growth scenario for 2026 plausible; the Banque de France forecasts growth of and Insee . For 2027 it finds the growth forecast optimistic, and the spending forecast attainable if every measure announced is carried out. Interest on the debt would come to in 2026 and to in 2027. The Government's figures in these rows are those of the bill as tabled: in its first sitting of 14 October 2026, the Assemblée nationale deleted the bill's article liminaire, which sets out the forecast balance, debt, levies and spending of general government for 2025 to 2027 (record of the sitting).