Theme

Official figures and this site's calculations

Who pays, who receives: tables by tenth of living standard, conventions and alternatives

Method working paper no. 1 of this site: how the tables of who pays and who receives by tenth of living standard are built for 2023 and 2024, the conventions they rest on and what changes under the alternatives.

Thomas Gazagnaire · How France works · version unreleased-ef7a7ba2 of 2026-11-12

Abstract

This paper describes how the site spreads 2023 and 2024 public levies and transfers over the tenths of living standard, and what changes when a convention changes. For 2023 the split is Insee's distributional national accounts (Insee Analyses no. 118) as published; for 2024, each of Insee's rows grows by its own national-accounts transaction, each group keeping its share. In 2023, each person of the poorest tenth thus pays €400 a month, estimated on average and receives €1,570, estimated; in the richest tenth, each pays €5,340, estimated and receives €1,990, estimated.

These amounts rest on conventions that Insee chose and the site keeps. Pensions are transfers, and collective spending is spread over people. VAT falls on consumers, and corporate tax on the owners of companies. Section 5 gives, for each, the published alternative and what the site computes of it today. When replacement income (pensions, unemployment benefits and sick pay) counts as deferred income, each person of the middle tenth pays €790 a month, estimated instead of €1,120, estimated and receives €1,050, estimated instead of €1,620, estimated. Section 6 reconciles the tables with Insee's published figures, section 7 with the IPP figures the site cites, and section 8 sets the site's concepts beside those of Insee, OpenFisca and TAXIPP.

Estimated: a model, an imputation or a surveyUnmarked figures are published as observed.

Section 1

Purpose and scope

The “Who pays, who receives” block of the Public money page attributes every public levy and transfer to groups of people ranked by their household's living standard. This paper sets out its method for two years, 2023 and 2024, and for the view by tenth, the reference one. The views by share of income and by age use the same totals; they are described here only for their edges.

Two fields overlap. The ranking comes from the survey on tax and social incomes (ERFS), whose field is metropolitan France, households in ordinary dwellings whose reference person is not a student. The amounts are those of the national accounts of all of France, and Insee counts 68.4 million, estimated people in its 2023 distributional accounts. The survey's field leaves out the residents of the overseas departments (2.2 million, estimated on 1 January 2024), people living in communities (1.6 million, estimated in 2019) and homeless people (141,500, estimated in metropolitan France early in 2012, Insee's last count). The site does not have the rule by which Insee places them in a twentieth.

The site's unit is the person. A group's amount, in billions of euros, divided by its people and by twelve, gives an amount a month per person, children included, rounded to 10 euros. Insee publishes its figures in euros per consumption unit (CU) and counts 47.0 million, estimated CU for everyone: an amount per CU is the amount per person times the group's ratio of people to CU. The site also publishes its amounts per CU, on the “Where do you stand?” page and in the tables by tenth. It then divides by the CU Insee's figures imply, the tenth's living standard in billions (Figure 2b) over the same per CU (Figure 2a): 4.7, estimated million CU for the middle tenth. Figure 2a prints that count to 0.1 million, and the sum of the twentieths, each rounded, is up to 2% off. Each tenth holds a tenth of Insee's people, and a person living alone counts one CU: an amount per CU is the tenth's amount over its units, not one person's bill.

Section 2

Sources

Each source is read by a program of the repository from a file kept under data/income/raw/, and each figure opens the panel that names its cell. The sources are these.

  • André M., Barrau A. and Renaud T., “Revenus des ménages élargis à l'ensemble de l'économie en 2023”, Insee Analyses no. 118, 16 April 2026, workbook of the figures: the supplementary table by twentieth, Figures 2a (per consumption unit) and 2b by tenth and Figure 1e by household age, extracted into income/dna_2023.csv. Everyone's extended primary income there is €2,083bn, estimated.
  • André M. and Renaud T., “Comptes nationaux par catégories de ménages en 2023 : note méthodologique”, Insee, version 2.0 of 16 April 2026: each row of the detailed table is a national-accounts transaction of the whole economy (p. 3), the rule for collective spending is on p. 12 and the bridge from the balance to the deficit on p. 13. Insee, table CND.101 (T_CND_101.xlsx, 16 April 2026): the same accounts for 2020 to 2023, with the same method, extracted into income/cnd_2020_2023.csv.
  • Insee, “Distribution des niveaux de vie, données annuelles de 1996 à 2024” (ERFS): the deciles and the 95th percentile of living standard, in constant 2024 euros, extracted into living_standards.csv. The 2023 median there is €26,280 a year, estimated.
  • Insee, “Les revenus et le patrimoine des ménages”, 2024 edition, sheet “Très hauts revenus” (Filosofi 2021): the thresholds of P99, P99.9 and P99.99, extracted into top_living_standards.csv. The P99 threshold there is €90,140.
  • Eurostat: the sector accounts by transaction (nasa_10_nf_tr), which grow each of Insee's rows, general government accounts by transaction (gov_10a_main) and by function (gov_10a_exp, COFOG), debt by holder (gov_10dd_ggd), shares by holding sector (nasa_10_f_bs), household consumption (nama_10_fcs) and population (demo_pjan). 2023 public revenue there is €1,456bn and spending €1,608bn.
  • The Commission des comptes de la Sécurité sociale, for the CSG's yield by base: capital income is 11.3%, estimated of it in 2023. The Banque de France, for the share of the CAC 40 companies held by non-residents (49.5%).
  • Two cross-checks, which enter no group: the World Inequality Database (WID), for the shape of the top of the distribution (P99–100 holds 48.4%, estimated of the top 5%'s income there), and IPP Note 92, for tax rates at the very top (section 7).

Section 3

The ranking and the groups

Insee ranks individuals by their household's living standard, as the survey measures it, and cuts them into twenty groups of 5% each. The poorest tenth joins the first two twentieths, and so on; the richest tenth joins the nineteenth and the twentieth. The site re-ranks no one: each group keeps the share Insee gives it of every levy and every transfer.

The edges in euros do not come from IA118, which publishes none. For the tenths, they are the survey's deciles of living standard for the year, in constant 2024 euros, divided by twelve and rounded to 10 euros: in 2023 the median living standard is €2,190 a month, estimated. The distributional accounts rank people by a living standard close to the survey's but not identical to it, so a person near an edge may belong to the next tenth. Table 1 gives each tenth's edges, its people, what each person pays and receives a month on average, and the share of its people who receive more than they pay, which Insee publishes (Figure 2a, row 26).

Table 1. The tenths of living standard: edges (constant 2024 euros a month, one person living alone), people in 2023, what each person, children included, pays and receives a month in current euros, and the share of net beneficiaries in 2023

Table 1. The tenths of living standard: edges (constant 2024 euros a month, one person living alone), people in 2023, what each person, children included, pays and receives a month in current euros, and the share of net beneficiaries in 2023
TenthPeoplePays, 2023Receives, 2023Pays, 2024Receives, 2024Net beneficiaries
Bottom 10%D1 · 6.8 million, estimated peopleunder €1,140, estimated, mean €840 a month, estimated6.8 million, estimated€400, estimated€1,570, estimated€410, estimated€1,620, estimated98.6%, estimated
Bottom 10–20%D2 · 6.8 million, estimated people€1,140, estimated to €1,440 a month, estimated6.8 million, estimated€550, estimated€1,820, estimated€560, estimated€1,870, estimated96.6%, estimated
Bottom 20–30%D3 · 6.8 million, estimated people€1,440, estimated to €1,720 a month, estimated6.8 million, estimated€740, estimated€1,740, estimated€760, estimated€1,790, estimated86.6%, estimated
Bottom 30–40%D4 · 6.8 million, estimated people€1,720, estimated to €1,960 a month, estimated6.8 million, estimated€940, estimated€1,740, estimated€970, estimated€1,800, estimated73.4%, estimated
Bottom 40–50%D5 · 6.8 million, estimated people€1,960, estimated to €2,190 a month, estimated6.8 million, estimated€1,120, estimated€1,620, estimated€1,150, estimated€1,680, estimated53.4%, estimated
Top 40–50%D6 · 6.8 million, estimated people€2,190, estimated to €2,440 a month, estimated6.8 million, estimated€1,300, estimated€1,660, estimated€1,340, estimated€1,710, estimated44.8%, estimated
Top 30–40%D7 · 6.8 million, estimated people€2,440, estimated to €2,740 a month, estimated6.8 million, estimated€1,560, estimated€1,670, estimated€1,610, estimated€1,720, estimated35.8%, estimated
Top 20–30%D8 · 6.8 million, estimated people€2,740, estimated to €3,180 a month, estimated6.8 million, estimated€1,920, estimated€1,620, estimated€1,980, estimated€1,670, estimated28.8%, estimated
Top 10–20%D9 · 6.8 million, estimated people€3,180, estimated to €3,990 a month, estimated6.8 million, estimated€2,450, estimated€1,720, estimated€2,520, estimated€1,780, estimated24.9%, estimated
Top 10%D10 · 6.8 million, estimated peopleover €3,990, estimated, mean €6,170 a month, estimated6.8 million, estimated€5,340, estimated€1,990, estimated€5,500, estimated€2,070, estimated18.2%, estimated

Above the 95th percentile the survey publishes no threshold, and the view by share of income needs one to bound its last groups. The site takes Filosofi's 2021 thresholds (tax and benefit files) at P99, P99.9 and P99.99, divides them by Filosofi's ninth decile and multiplies them by the survey's ninth decile of the year: Filosofi gives the shape of the top of the distribution, the survey its level. Filosofi's ninth decile is €41,220 in 2021, which puts P99 at €104,748 a year, estimated in 2023. Between two published thresholds above P95, the living standard follows a Pareto segment: its logarithm is linear in the logarithm of the share of people above it. Below P95 it is linear between two deciles. No edge of the page lies above P99.99, so nothing is extrapolated.

The view by share of income cuts the same ranking into groups that each hold the same share of extended primary income: nine groups of 10% of that income, then five of 2%. The rank edges are read on Insee's concentration curve, linear between its twentieths below P95 and, above, on a generalized Pareto curve (Blanchet, Fournier and Piketty, 2022) fitted on Insee's twenty points. Half of extended primary income thus goes to the people above percentile 79.7, estimated. The curve's inverted Pareto coefficient is 1.9, estimated at P95 and its shape parameter 0.5, estimated; on it, P99–100 holds 43.5%, estimated of the top 5%'s income, against 48.4%, estimated in WID, which ranks adults by their own income. The gap goes the expected way: ranking people by their household's living standard rather than their own income can only lower the top's share. Table 2 gives these groups.

Table 2. The groups by share of income, 2023: edges, people, and what each person pays and receives a month, with in brackets the lowest and highest of the four variants

Table 2. The groups by share of income, 2023: edges, people, and what each person pays and receives a month, with in brackets the lowest and highest of the four variants
GroupPeoplePaysReceives
Bottom 33%P0–33 · 23 million, estimated peopleunder €1,800 a month, estimated22.8 million, estimated€590, estimated€1,710, estimated
Bottom 33–49%P33–49 · 11 million, estimated people€1,800, estimated to €2,170 a month, estimated10.7 million, estimated€1,050, estimated€1,670, estimated
Middle, 49–61% from the bottomP49–61 · 8.5 million, estimated people€2,170, estimated to €2,480 a month, estimated8.5 million, estimated€1,320, estimated€1,650, estimated
Top 29–39%P61–71 · 6.9 million, estimated people€2,480, estimated to €2,810 a month, estimated6.9 million, estimated€1,600, estimated€1,660, estimated
Top 20–29%P71–80 · 5.7 million, estimated people€2,810, estimated to €3,170 a month, estimated5.7 million, estimated€1,930, estimated€1,620, estimated
Top 13–20%P80–87 · 4.7 million, estimated people€3,170, estimated to €3,720 a month, estimated4.7 million, estimated€2,320, estimated€1,690, estimated
Top 7.9–13%P87–92 · 3.8 million, estimated people€3,720, estimated to €4,420 a month, estimated3.8 million, estimated€2,920, estimated€1,810, estimated
Top 3.5–7.9%P92–96.5 · 3 million, estimated people€4,420, estimated to €5,600 a month, estimated3.0 million, estimated€3,600, estimated [€3,530, estimated–€3,830, estimated]€1,910, estimated [€1,890, estimated–€1,910, estimated]
Top 0.9–3.5%P96.5–99.1 · 1.8 million, estimated people€5,600, estimated to €9,000 a month, estimated1.8 million, estimated€5,910, estimated [€5,210, estimated–€6,090, estimated]€2,080, estimated [€2,060, estimated–€2,080, estimated]
Top 0.6–0.9%P99.1–99.4 · 220,000, estimated people€9,000, estimated to €10,800 a month, estimated221,681, estimated€9,490, estimated [€7,750, estimated–€9,540, estimated]€2,200, estimated [€2,170, estimated–€2,290, estimated]
Top 0.3–0.6%P99.4–99.7 · 180,000, estimated people€10,800, estimated to €13,600 a month, estimated175,047, estimated€12,010, estimated [€9,560, estimated–€12,060, estimated]€2,280, estimated [€2,250, estimated–€2,400, estimated]
Top 0.2–0.3%P99.7–99.8 · 130,000, estimated people€13,600, estimated to €18,900 a month, estimated127,358, estimated€16,510, estimated [€12,900, estimated–€16,520, estimated]€2,430, estimated [€2,410, estimated–€2,580, estimated]
Top 0.04–0.2%P99.8–99.96 · 78,000, estimated people€18,900, estimated to €38,000 a month, estimated78,339, estimated€26,840, estimated [€21,260, estimated–€28,490, estimated]€2,780, estimated [€2,770, estimated–€2,920, estimated]
Top 0.04%P99.96–100 · 27,000, estimated peopleover €38,000 a month, estimated27,208, estimated€77,290, estimated [€74,830, estimated–€192,930, estimated]€4,460, estimated [€3,940, estimated–€6,030, estimated]

Section 4

Insee's accounts and the years after 2023

Insee publishes its distributional accounts by twentieth for 2020 to 2023, with one method (table CND.101), and IA118 details 2023. Each row of the detailed table is a national-accounts transaction of the whole economy (CND note, p. 3): taxes on products are D21 received by general government, households' taxes on income the D5 they pay, contributions the D61 they pay, cash benefits the D62 and transfers in kind the D63 they receive, collective spending general government's collective consumption (P32). Eurostat's sector accounts give each one back to €0.1bn (Table 3). For 2023, the site takes Insee's cells as they are.

Table 3. Insee's rows, the national-accounts transaction each one is, its 2023 amount in Insee's accounts and Eurostat's, and its growth in 2024

Table 3. Insee's rows, the national-accounts transaction each one is, its 2023 amount in Insee's accounts and Eurostat's, and its growth in 2024
TransactionInsee's rows, 2023Eurostat, 2023Eurostat, 2024Growth, 2024
Taxes on products received by general government (D21)€318.6bn, estimated€318.6bn€324.7bn1.02, estimated
Subsidies on products paid by general government (D31)€31.8bn, estimated€31.8bn€17.1bn0.54, estimated
Other taxes on production received by general government (D29)€128bn, estimated€128bn€129bn1.01, estimated
Other subsidies on production paid by general government (D39)€38.1bn, estimated€38.1bn€40.6bn1.06, estimated
Taxes on income paid by corporations (D5, S11 and S12)€78bn, estimated€78bn€80.1bn1.03, estimated
Taxes on income and wealth paid by households (D5, S14)€273.2bn, estimated€273.3bn€280.7bn1.03, estimated
Social contributions paid by households (D61, S14)€542.9bn, estimated€542.9bn€571.8bn1.05, estimated
Cash social benefits received by households (D62, S14)€591.7bn, estimated€591.8bn€631.2bn1.07, estimated
Social transfers in kind received by households (D63, S14)€505.9bn, estimated€505.9bn€524.3bn1.04, estimated
Collective consumption of general government (P32)€241.5bn, estimated€241.5bn€249.5bn1.03, estimated

The perimeter is Insee's, the whole economy: besides general government, the pension and health schemes of employers and insurers and the services of non-profit institutions. The diagram still keeps Eurostat's revenue and spending of general government. The contributions the other schemes receive (€81.3bn in 2023) and the benefits and services they pay (€127.4bn) are drawn apart. What Eurostat counts and Insee does not attribute stays in view without being spread: other revenue (sales, fees, property income) is €164.5bn, estimated, inheritance and gift duties €20.8bn and the spending Insee does not attribute, interest aside, €242.2bn, estimated (investment, capital transfers and other transfers, section 5.4). Borrowing is the year's deficit, €151.9bn.

For 2024, no one publishes a split between groups. Each of Insee's rows grows by its own transaction, the ratio of 2024 to 2023 read in one vintage of the sector accounts so that a revision of the 2023 level never enters; each group keeps its share of the row, and row 48 (balances with the rest of the world) is held at its level. People and consumption units grow with the population, 68.8 million in 2024 against 68.5 million in 2023, as the WID guidelines recommend (Blanchet et al., 2021, §8.2, p. 154). Each person of the middle tenth (D5) thus pays €1,150 a month, estimated in 2024, against €1,120, estimated in 2023.

The assumption is shares unchanged and totals right. Insee's series test it: carrying 2022's shares to 2023's totals, row by row, misses Insee's net transfer by €151 a month per CU, estimated for the poorest tenth, €71, estimated for the middle tenth and €−415, estimated for the richest (a positive gap gives the tenth too much). The 2024 figures carry an uncertainty of that order, on top of that of Insee's own accounts.

Table 3b reconciles the 2023 figures with Insee's Figure 2a, per consumption unit. The site divides the cells of the detailed table by the units Insee's figures imply (section 1); the gap, €150 a year at most, is rounding. The last column gives the net transfer of the method the site used until October 2026: each Insee item scaled to its category's Eurostat total, general government's receipts or total expenditure by function (COFOG). Total expenditure also counts gross investment, capital transfers and the EU contribution, which no household receives in the year, and general government's receipts leave out the other schemes' contributions; the method thus put each tenth €200 to €800 a month per CU away from Insee, and it is retired.

Table 3b. Per consumption unit a year, 2023: Insee's Figure 2a (rows 10, 14, 17 and 24) and the site, which takes the cells of the detailed table, with the net transfer of the previous method

Table 3b. Per consumption unit a year, 2023: Insee's Figure 2a (rows 10, 14, 17 and 24) and the site, which takes the cells of the detailed table, with the net transfer of the previous method
TenthLevies (i), InseesiteCash (ii), InseesiteNon-cash (iii), InseesiteNet transfer, InseesitePrevious method
Bottom 10%D1 · 6.8 million, estimated peopleunder €1,140, estimated, mean €840 a month, estimated€6,400, estimated€6,321, estimated€8,000, estimated€8,025, estimated€19,100, estimated€19,114, estimated€−20,800, estimated€−20,818, estimated€−23,529, estimated
Bottom 10–20%D2 · 6.8 million, estimated people€1,140, estimated to €1,440 a month, estimated€8,800, estimated€8,761, estimated€10,600, estimated€10,603, estimated€20,500, estimated€20,477, estimated€−22,300, estimated€−22,319, estimated€−25,336, estimated
Bottom 20–30%D3 · 6.8 million, estimated people€1,440, estimated to €1,720 a month, estimated€11,900, estimated€11,893, estimated€11,700, estimated€11,703, estimated€17,400, estimated€17,406, estimated€−17,200, estimated€−17,216, estimated€−20,307, estimated
Bottom 30–40%D4 · 6.8 million, estimated people€1,720, estimated to €1,960 a month, estimated€15,100, estimated€15,065, estimated€11,500, estimated€11,534, estimated€17,200, estimated€17,217, estimated€−13,700, estimated€−13,686, estimated€−17,150, estimated
Bottom 40–50%D5 · 6.8 million, estimated people€1,960, estimated to €2,190 a month, estimated€18,200, estimated€18,283, estimated€12,000, estimated€12,005, estimated€15,000, estimated€15,017, estimated€−8,800, estimated€−8,739, estimated€−12,153, estimated
Top 40–50%D6 · 6.8 million, estimated people€2,190, estimated to €2,440 a month, estimated€21,500, estimated€21,377, estimated€12,600, estimated€12,558, estimated€15,000, estimated€15,002, estimated€−6,100, estimated€−6,184, estimated€−9,868, estimated
Top 30–40%D7 · 6.8 million, estimated people€2,440, estimated to €2,740 a month, estimated€25,700, estimated€25,716, estimated€12,800, estimated€12,784, estimated€14,700, estimated€14,692, estimated€−1,800, estimated€−1,760, estimated€−5,848, estimated
Top 20–30%D8 · 6.8 million, estimated people€2,740, estimated to €3,180 a month, estimated€31,700, estimated€31,692, estimated€13,400, estimated€13,394, estimated€13,000, estimated€13,012, estimated€5,300, estimated€5,286, estimated€992, estimated
Top 10–20%D9 · 6.8 million, estimated people€3,180, estimated to €3,990 a month, estimated€41,100, estimated€41,078, estimated€14,600, estimated€14,602, estimated€13,600, estimated€13,635, estimated€12,900, estimated€12,841, estimated€7,717, estimated
Top 10%D10 · 6.8 million, estimated peopleover €3,990, estimated, mean €6,170 a month, estimated€89,800, estimated€89,746, estimated€18,500, estimated€18,433, estimated€12,900, estimated€12,912, estimated€58,400, estimated€58,401, estimated€51,389, estimated

Section 5

Conventions and their alternatives

Each convention below is Insee's, which the site keeps. For each, the text gives the published alternative and what the site computes of it today; where it does not compute it, it says so.

5.1 Replacement income

Insee counts retirement pensions, unemployment benefits and sick pay as public transfers, and the contributions that fund them as levies. Its box “Traitement des revenus de remplacement” publishes the reverse treatment: that income, “les pensions de retraite, les allocations chômage ou encore les indemnités journalières”, becomes primary income, and the levies that fund it, “notamment les cotisations retraite et chômage”, are neutralised. The ratio of the richest tenth's mean income before transfers to the poorest tenth's then falls from 26, estimated to 14, estimated; income after transfers does not change. Insee's glossary of the household accounts defines replacement income as retirement pensions, unemployment benefits, sick pay and work-injury and occupational-disease annuities: rows 31, 32 and 33 of the detailed table. Insee does not publish this treatment by tenth.

The site computes it by tenth, on the box's items: rows 31 to 33 leave what each tenth receives, pension and unemployment contributions (rows 26 and 27) what it pays, and the CSG on that income stays a levy. Table 4 gives those items for 2023, beside all the tenth pays and receives. Pensions are €10.6bn, estimated of the €129.1bn, estimated the poorest tenth receives, and €66.6bn, estimated of the richest tenth's €163.5bn, estimated. A month per person, the middle tenth (D5) then pays €790, estimated instead of €1,120, estimated and receives €1,050, estimated instead of €1,620, estimated; the poorest tenth receives €1,370, estimated instead of €1,570, estimated, the richest €990, estimated instead of €1,990, estimated. Income after transfers does not change: what leaves a tenth's balance enters its income before transfers. The groups stay Insee's, ranked by living standard, which the convention does not change, and the gap between the benefits moved and the contributions neutralised is not rebalanced, as in the box.

On Insee's cells, per consumption unit, the ratio of the richest tenth to the poorest is then 17.27, estimated on extended primary income, the base of Insee's 26: that is the site's figure. It is 14.36, estimated on households' primary income, without retained profits and public bodies' primary income; on that base the cells give back Insee's 14. The box does not name its base, and no set of rows gives 14 on extended primary income: Insee most likely computes its 14 on households' primary income, an inference only Insee can confirm.

Two other conventions are published. Bozio, Garbinti, Goupille-Lebret, Guillot and Piketty (2020, pp. 7-8 and note 17, p. 10) deduct pension and unemployment contributions and add pensions and unemployment benefits, without sick pay: the site keeps it beside the switch, and the middle tenth receives €1,090, estimated a month per person under it. The “broad” definition of the WID guidelines (Blanchet et al., 2021, §2.2.3.1, p. 61) deducts all social contributions and adds all social insurance benefits, family benefits aside; the site does not compute it. The three differ by less than one ratio point on either base.

Table 4. Replacement income and the contributions that fund it by tenth, and what the tenth pays and receives when it counts as deferred income, 2023, €bn

Table 4. Replacement income and the contributions that fund it by tenth, and what the tenth pays and receives when it counts as deferred income, 2023, €bn
TenthPension contributionsUnemployment contributionsAll it paysPays, deferred incomePensionsUnemployment benefitsSick pay and work-injury annuities (row 33)All it receivesReceives, deferred income
Bottom 10%D1 · 6.8 million, estimated peopleunder €1,140, estimated, mean €840 a month, estimated€3.9bn, estimated€300m, estimated€32.6bn, estimated€28.4bn, estimated€10.6bn, estimated€5.8bn, estimated€600m, estimated€129.1bn, estimated€112.1bn, estimated
Bottom 10–20%D2 · 6.8 million, estimated people€1,140, estimated to €1,440 a month, estimated€9.8bn, estimated€900m, estimated€44.8bn, estimated€34.1bn, estimated€22.7bn, estimated€5.7bn, estimated€1.3bn, estimated€149bn, estimated€119.3bn, estimated
Bottom 20–30%D3 · 6.8 million, estimated people€1,440, estimated to €1,720 a month, estimated€14.7bn, estimated€1.5bn, estimated€61bn, estimated€44.8bn, estimated€34.2bn, estimated€5.1bn, estimated€1.9bn, estimated€142.5bn, estimated€101.3bn, estimated
Bottom 30–40%D4 · 6.8 million, estimated people€1,720, estimated to €1,960 a month, estimated€20.2bn, estimated€2bn, estimated€77.5bn, estimated€55.3bn, estimated€36.3bn, estimated€4.8bn, estimated€2.6bn, estimated€143bn, estimated€99.3bn, estimated
Bottom 40–50%D5 · 6.8 million, estimated people€1,960, estimated to €2,190 a month, estimated€24.9bn, estimated€2.5bn, estimated€91.9bn, estimated€64.5bn, estimated€39bn, estimated€4.6bn, estimated€3.1bn, estimated€133.1bn, estimated€86.4bn, estimated
Top 40–50%D6 · 6.8 million, estimated people€2,190, estimated to €2,440 a month, estimated€28.4bn, estimated€2.9bn, estimated€107bn, estimated€75.7bn, estimated€43.6bn, estimated€3.5bn, estimated€3.5bn, estimated€136.1bn, estimated€85.5bn, estimated
Top 30–40%D7 · 6.8 million, estimated people€2,440, estimated to €2,740 a month, estimated€35.6bn, estimated€3.8bn, estimated€128.4bn, estimated€89bn, estimated€44.7bn, estimated€3.8bn, estimated€4.1bn, estimated€136.7bn, estimated€84.1bn, estimated
Top 20–30%D8 · 6.8 million, estimated people€2,740, estimated to €3,180 a month, estimated€43.5bn, estimated€4.7bn, estimated€157.5bn, estimated€109.3bn, estimated€47.8bn, estimated€3.4bn, estimated€5bn, estimated€132.6bn, estimated€76.4bn, estimated
Top 10–20%D9 · 6.8 million, estimated people€3,180, estimated to €3,990 a month, estimated€54.7bn, estimated€5.8bn, estimated€200.9bn, estimated€140.4bn, estimated€52.3bn, estimated€3.7bn, estimated€6bn, estimated€141.1bn, estimated€79.1bn, estimated
Top 10%D10 · 6.8 million, estimated peopleover €3,990, estimated, mean €6,170 a month, estimated€93bn, estimated€8.8bn, estimated€438.5bn, estimated€336.7bn, estimated€66.6bn, estimated€5.4bn, estimated€9.9bn, estimated€163.5bn, estimated€81.6bn, estimated

5.2 Collective services

Collective spending (defence, public order, justice, general administration, economic affairs) and culture go to no one in particular, and Insee spreads them over people. For culture and leisure the key is an equal amount per person: €765 a year per person, estimated in the first twentieth as in the last (€765, estimated). For collective spending, Insee's cells do not give an equal amount per person: €3,382 a year, estimated in the tenth twentieth, €4,118, estimated in the first and €4,235, estimated in the last. The methodological note of the household accounts (André and Renaud, version 2.0, April 2026, p. 12) gives the rule: collective spending of national attribution, such as defence and foreign affairs, is spread evenly over individuals; the main part, called localisable (police and gendarmerie, justice, general administration), by the pay bill of the public staff concerned over the inhabitants of each household's bassin de vie. The note finds a U-shaped profile by living standard. The site takes the cells as they are, and its panels say they are not equal per person. Table 5 gives the 2023 amounts by tenth, with the people of each tenth.

Table 5. Collective services and culture by tenth, the people of each tenth, and what it receives under each convention, 2023, €bn

Table 5. Collective services and culture by tenth, the people of each tenth, and what it receives under each convention, 2023, €bn
TenthPeopleCollective services, InseeCollective services by incomeCulture and leisureReceives, InseeReceives, by incomeReceives, services left out
Bottom 10%D1 · 6.8 million, estimated peopleunder €1,140, estimated, mean €840 a month, estimated6.8 million, estimated€26.5bn, estimated€5.9bn, estimated€5.2bn, estimated€129.1bn, estimated€108.5bn, estimated€102.6bn, estimated
Bottom 10–20%D2 · 6.8 million, estimated people€1,140, estimated to €1,440 a month, estimated6.8 million, estimated€24.6bn, estimated€11.3bn, estimated€5.2bn, estimated€149bn, estimated€135.7bn, estimated€124.4bn, estimated
Bottom 20–30%D3 · 6.8 million, estimated people€1,440, estimated to €1,720 a month, estimated6.8 million, estimated€24.2bn, estimated€14.8bn, estimated€5.2bn, estimated€142.5bn, estimated€133.1bn, estimated€118.3bn, estimated
Bottom 30–40%D4 · 6.8 million, estimated people€1,720, estimated to €1,960 a month, estimated6.8 million, estimated€24bn, estimated€17.2bn, estimated€5.2bn, estimated€143bn, estimated€136.2bn, estimated€119bn, estimated
Bottom 40–50%D5 · 6.8 million, estimated people€1,960, estimated to €2,190 a month, estimated6.8 million, estimated€23.4bn, estimated€19.2bn, estimated€5.2bn, estimated€133.1bn, estimated€128.9bn, estimated€109.7bn, estimated
Top 40–50%D6 · 6.8 million, estimated people€2,190, estimated to €2,440 a month, estimated6.8 million, estimated€23.4bn, estimated€21.3bn, estimated€5.2bn, estimated€136.1bn, estimated€134bn, estimated€112.7bn, estimated
Top 30–40%D7 · 6.8 million, estimated people€2,440, estimated to €2,740 a month, estimated6.8 million, estimated€23.4bn, estimated€24bn, estimated€5.2bn, estimated€136.7bn, estimated€137.3bn, estimated€113.3bn, estimated
Top 20–30%D8 · 6.8 million, estimated people€2,740, estimated to €3,180 a month, estimated6.8 million, estimated€23bn, estimated€27.7bn, estimated€5.2bn, estimated€132.6bn, estimated€137.3bn, estimated€109.6bn, estimated
Top 10–20%D9 · 6.8 million, estimated people€3,180, estimated to €3,990 a month, estimated6.8 million, estimated€23.2bn, estimated€33bn, estimated€5.2bn, estimated€141.1bn, estimated€150.9bn, estimated€117.9bn, estimated
Top 10%D10 · 6.8 million, estimated peopleover €3,990, estimated, mean €6,170 a month, estimated6.8 million, estimated€23.8bn, estimated€65.2bn, estimated€5.2bn, estimated€163.5bn, estimated€204.9bn, estimated€139.7bn, estimated

Two other conventions are published, and the site computes them by tenth. The distributional national accounts of Piketty, Saez and Zucman (Quarterly Journal of Economics, 2018) spread collective consumption in proportion to income; for France, Bozio, Garbinti, Goupille-Lebret, Guillot and Piketty (2020, p. 18) take income after levies and cash benefits, and the site does the same with Insee's rows 14, 17 and 30. The national accounts' adjusted disposable income, which adds only the services each person uses (System of National Accounts 2008, §8.32), leaves that consumption out; the site then takes it out of what each tenth receives. Both conventions move Insee's row 46, collective spending and balances with the rest of the world; culture keeps Insee's split, and what each tenth pays does not change. The last three columns of Table 5 give the amounts. A month per person, the poorest tenth receives €1,320, estimated in proportion to income and €1,250, estimated without these services, instead of €1,570, estimated; the richest €2,500, estimated and €1,700, estimated instead of €1,990, estimated. The first convention lowers what the poorest tenths receive and raises what the richest receive; the second lowers what every tenth receives.

5.3 Incidence of VAT, employer contributions and corporate tax

Three families of levies are paid by companies and attributed to households by convention. Insee attributes taxes on products (VAT, excise duties) to the households who consume, taxes on production and subsidies to households in proportion to what they draw from production, and corporate tax to the owners of companies. Social contributions, the employer's part included, count as part of pay and fall on employees; Insee's table does not separate employer from employee contributions. Table 6 gives, by tenth, the levies that pass through companies, beside all the tenth pays.

Table 6. Levies that pass through companies, by tenth, 2023, €bn

The site computes two alternatives. The first, a variant the diagram does not draw, concerns what would fall outside France (section 5.5). Non-residents hold 20%, estimated of the shares of French companies in the financial accounts, which puts €15.6bn, estimated of corporate tax outside France in 2023, or €38.7bn, estimated at the CAC 40's rate of foreign holding (49.5%); the two bound the true share. The second concerns the top of the distribution, where Insee publishes one group of 5%. The central case spreads every levy there in proportion to income, an assumption of the site that overstates the VAT paid at the very top. Four variants change one assumption at a time, and Table 2 gives their range in brackets. The first takes a strict Pareto tail, the second the most concentrated tail compatible with WID. The third spreads each item along Insee's gradient between its last two twentieths; the fourth spreads personal taxes, contributions and corporate tax along IPP's rates. In the last group, what each person pays a month runs from €74,830, estimated to €192,930, estimated, against €77,290, estimated in the central case. The site computes no other incidence of employer contributions, nor a part of corporate tax that would fall on wages.

5.4 Debt interest and borrowing

Levies do not cover all spending, and the general-government deficit, €151.9bn in 2023, finances the rest. Insee attributes that balance to the groups as a future levy (row “solde de la redistribution publique nationale”, €66.2bn, estimated on its perimeter), rising with living standard (Table 7). The site does not take up that split. It draws borrowing to the lenders who took up the year's debt, from the change in their holdings (Eurostat, debt by holder). In 2023, lenders outside France took up 104.2%, estimated, other lenders in France 0.4%, estimated and the Banque de France −4.6%, estimated; a negative share is debt that matured and that the State repaid with new borrowing.

Table 7. The balance Insee attributes to each tenth as a future levy, 2023, €bn

Table 7. The balance Insee attributes to each tenth as a future levy, 2023, €bn
TenthBalance attributed by InseeAll it pays
Bottom 10%D1 · 6.8 million, estimated peopleunder €1,140, estimated, mean €840 a month, estimated€3.8bn, estimated€32.6bn, estimated
Bottom 10–20%D2 · 6.8 million, estimated people€1,140, estimated to €1,440 a month, estimated€4.5bn, estimated€44.8bn, estimated
Bottom 20–30%D3 · 6.8 million, estimated people€1,440, estimated to €1,720 a month, estimated€4.8bn, estimated€61bn, estimated
Bottom 30–40%D4 · 6.8 million, estimated people€1,720, estimated to €1,960 a month, estimated€5.3bn, estimated€77.5bn, estimated
Bottom 40–50%D5 · 6.8 million, estimated people€1,960, estimated to €2,190 a month, estimated€5.4bn, estimated€91.9bn, estimated
Top 40–50%D6 · 6.8 million, estimated people€2,190, estimated to €2,440 a month, estimated€5.8bn, estimated€107bn, estimated
Top 30–40%D7 · 6.8 million, estimated people€2,440, estimated to €2,740 a month, estimated€6.4bn, estimated€128.4bn, estimated
Top 20–30%D8 · 6.8 million, estimated people€2,740, estimated to €3,180 a month, estimated€7bn, estimated€157.5bn, estimated
Top 10–20%D9 · 6.8 million, estimated people€3,180, estimated to €3,990 a month, estimated€8.3bn, estimated€200.9bn, estimated
Top 10%D10 · 6.8 million, estimated peopleover €3,990, estimated, mean €6,170 a month, estimated€14.9bn, estimated€438.5bn, estimated

Insee's balance is not the deficit. The CND note (p. 13) goes from one to the other through interest net of financial income, investment net of the consumption of fixed capital, capital transfers and the balance with the rest of the world. Eurostat's sector accounts give the lines back for 2023: to Insee's balance (€−66.2bn, estimated) add property income received less interest paid (€−33.3bn), net investment (€−11.8bn), net capital transfers (€−12.1bn) and the rest, the rest of the world included (€−28.3bn, estimated), for net borrowing of €−151.7bn. Insee gives −66, −33, −12, and −40 for the last two lines together.

Debt interest (€51.7bn in 2023) goes to no tenth. It is split between holders by their average holdings over the year: outside France 49.9%, estimated, the Banque de France 21.9%, estimated, other lenders in France 28.2%, estimated. What households receive of it through their insurers and funds is already in their primary income in Insee's accounts; the site does not count it twice.

5.5 Flows abroad

Insee attributes everything to residents: it takes the balance of French companies held from abroad and foreign companies held by residents as negligible (CND note, pp. 5 and 9). The diagram follows Insee. Its “Outside France” group carries only general government's flows with the rest of the world. In 2023 it pays €170.5bn, estimated: the loans of non-residents and the transfers received from the European Union. It receives €60.9bn, estimated: France's contribution to the Union (€23.9bn), international cooperation (€10.8bn), capital transfers (€373m) and the interest paid to non-resident holders, taken gross from the government accounts. The site's “Outside France” layer stays a variant: it would take from residents corporate tax at the rate of foreign holding (€15.6bn, estimated in 2023, section 5.3) and the part of taxes on products paid by visitors, their share of consumption on the territory (4.4%, estimated); it would put the richest tenth about €260 a month per CU away from Insee's Figure 2a.

Section 6

Reconciliation with Insee Analyses no. 118

What is the same. The ranking, the twenty groups, the list of items and each group's share of each item are Insee's. The view by tenth adds the cells of the supplementary table two by two, and each panel names the cells it adds.

What differs, and what agrees, in six points.

  • Subsidies. Insee publishes its levies net of subsidies: €1,341bn, estimated of levies less €69.9bn, estimated of subsidies give, to rounding, its total of €1,271bn, estimated. The site counts levies gross and puts subsidies among what the groups receive, so both sides are larger than Insee's.
  • The perimeter. The site keeps Insee's: it attributes €1,340bn, estimated of levies in 2023, Insee's gross levies above, and pensions are €397.8bn, estimated in the diagram as in Insee's accounts (€397.9bn, estimated, to the rounding of the twentieths). The schemes outside general government are drawn apart (section 4).
  • Per consumption unit. Divided by the units Insee's figures imply, the site's amounts give back Figure 2a to within €150 a year for every tenth (Table 3b).
  • The CSG on capital income. Insee does not split the CSG by base. The site takes out its part on capital income, from the yield by base of the Commission des comptes de la Sécurité sociale, and shares it by each group's financial income and rents, a key of the site. Each group's total CSG does not change.
  • Borrowing. Insee attributes its balance (€66.2bn, estimated) to the groups; the site draws the year's deficit (€151.9bn) to the lenders (section 5.4).
  • The unit and the year. The site publishes per person and per consumption unit, Insee per consumption unit (section 1); Insee publishes 2020 to 2023, the site adds 2024 by growing each row as its transaction, shares unchanged (section 4).

Section 7

Cross-checks against IPP figures

The site cites one IPP source for the split: Note 92 (Bach, Bozio, Guillouzouic and Malgouyres, updated May 2025), which gives for 2016 the tax rates of tax households by g-percentile of economic income, declared income plus the profits kept in the companies they control. It checks the top of the distribution and feeds one variant (section 5.3); it enters no group. Direct taxes, the corporate tax of controlled companies included, take on average 34.5%, estimated of the economic income of the richest 0.1%, 42%, estimated for P99–99.9, 46%, estimated at the entry to the top 0.1% and 26%, estimated for the 75 richest households. Personal taxes alone take 9.8%, estimated for the top 0.1% and 2%, estimated for the 378 richest households. Declared income is only 26.4%, estimated of the top 0.1%'s economic income.

In Insee's last twentieth, direct levies (income tax, CSG, other taxes on income and wealth, social contributions, corporate tax) take 49.5%, estimated of extended primary income in 2023. The two figures do not measure the same thing. The twentieth holds 5% of people ranked by living standard, IPP's slices are tax households ranked by economic income, far higher in the distribution. The year differs, 2023 against 2016, before the flat tax on capital income and the replacement of the wealth tax by a tax on property in 2018. Extended primary income also counts employer contributions and a share of public bodies' primary income. The gap is therefore no measure of either one's error.

What is missing. IPP publishes each year, with Cepremap, the budgetary and redistributive effects of tax and benefit measures by hundredth of living standard, computed with TAXIPP 2.4.2. Those results are not in the site, and no check by tenth against TAXIPP is made here. The two exercises answer different questions. TAXIPP measures the effect of new measures on direct levies and benefits; the site measures a level for the year, indirect levies and transfers in kind included.

Section 8

Concepts compared

Table 8 sets the site's concepts beside those of Insee, of OpenFisca-France and LexImpact, and of TAXIPP. Each cell sums up the public documentation read on 17 October 2026, listed below; “not verified” marks an inference the documentation does not state in words. OpenFisca's legislation explorer did not answer that day; the source code was read instead.

Table 8. The concepts compared

Table 8. The concepts compared
ConceptThis siteInsee, distributional accountsOpenFisca-France and LexImpactTAXIPP and IPP notes
RankingInsee's ranking, taken as it is; edges in euros from the ERFS and, above P95, from Filosofi.Usual living standard: the household's disposable income divided by its CU (IA118, the Ines model on the ERFS).OpenFisca: niveau_de_vie = revenu_disponible ÷ unites_consommation, at household level. LexImpact publishes results by decile; the concept behind the decile was not found.Perspectives budgétaires 2026: hundredths of initial living standard (disposable income per CU). Note 92: tax households by g-percentile of economic income.
UnitThe person: the group's amount a month per person, children included.Individuals are ranked; amounts are published per CU, on the modified OECD scale.Individual, family, tax household and household; unites_consommation follows Insee's scale.Individuals and CU, modified OECD scale (2026); tax households (Note 92).
PerimeterEvery levy and transfer Insee attributes, indirect, in-kind and collective ones included, on Insee's perimeter (the whole economy); flows abroad, schemes outside general government and borrowing drawn apart.Corporate tax, VAT and excise duties, contributions, benefits, transfers in kind and collective public services.Direct taxes, employee and employer contributions, benefits; VAT in the openfisca-france-indirect-taxation extension. LexImpact: income tax, CSG, contributions and benefits, no VAT found.TAXIPP 2.4.2: direct levies, not indirect ones. Guide 1.1 leaves out company taxes and employer contributions.
IncidenceInsee's, less the part that falls outside France (section 5.5).VAT by consumption, corporate tax on all owners, contributions on households whether levied on the employee or the employer (methodological note, 2024).No incidence assumption found: the model computes the amounts due by law (not verified).Guide 0.3: employer contributions paid in the long run by employees, indirect taxes shared between consumer prices and factor prices, corporate tax on capital income. Guide 1.1: no incidence of company taxes on households. Note 92: corporate tax to the holders of the capital.
PensionsTransfers, pension contributions counted as levies; as a variant by tenth, pensions, unemployment benefits and their contributions taken out of both sides (section 5.1).Transfers; in a box, replacement income moved into primary income and pension and unemployment contributions neutralised.pensions_nettes enters revenu_disponible; pension contributions reduce net pay (not verified).Guide 0.3: replacement transfers. 2026: pensions inside disposable income (not verified). Note 92: pension and unemployment contributions left out, pensions inside taxable income.

Sources of Table 8: Insee, IA118 · Insee, note méthodologique CND v1.1 · OpenFisca-France, mesures.py · LexImpact, préparation des données · TAXIPP 2.4.2 · TAXIPP, guide 0.3 · TAXIPP, guide 1.1 · IPP, Perspectives budgétaires 2026, ch. 2.

Section 9

Limits

The limits below belong to the method; those of Insee's accounts themselves are described in IA118.

  • A group's mean is not a person. In each tenth some receive more than they pay and others less; Insee publishes the share of net beneficiaries by tenth, which Table 1 gives, from 98.6%, estimated in the first to 18.2%, estimated in the last.
  • A year is not a lifetime. The tables describe one year; they do not say what a person pays and receives over a lifetime, and age weighs on them more than living standard.
  • 2024 is an extrapolation: each group keeps its 2023 share of every row, and its people are 2023's carried with the population. The test on Insee's own years gives the order of the error (section 4).
  • The edges come from the ERFS survey and, above P95, from Filosofi 2021 scaled to the survey. The distributional accounts rank people by a living standard close to the survey's, not identical to it.
  • Behaviour does not change: an alternative moves amounts with behaviour held fixed.
  • Some keys do not exist. Inheritance duties come from no group. The IFI and the property tax stay inside Insee's rows, which join them with other taxes. Taxes on production are not spread outside France.
  • The alternatives of sections 5.1 and 5.2 are computed by tenth, but only on the Insee rows they move: deferred income covers pensions, unemployment benefits and their contributions, not other replacement income. The variants of section 5.3 concern the top of the distribution only. No other incidence of VAT, employer contributions or corporate tax is computed: Insee's table does not separate employer from employee contributions, and putting VAT or corporate tax on factor incomes would need a share for each factor, which the site does not set.

Section 10

Version and citation

This paper is built with each version of the site from the same figures as its pages: its numbers cannot drift from those of the Public money page. This version is that of commit ef7a7ba2, of 2026-11-12. Corrections are logged on the Method page.

To cite this paper: Gazagnaire T. (2026), “Who pays, who receives: tables by tenth of living standard, conventions and alternatives”, How France works, working paper no. 1, version unreleased-ef7a7ba2, Licence Ouverte 2.0, https://patte-blanche.org/paper-1.html. The French text, https://patte-blanche.org/fr/paper-1.html, is the reference one.

The data: income/dna_2023.csv, income/living_standards.csv, income/top_living_standards.csv, income/SOURCES.md and claims.json, which gives each figure with its source and calculation.

Sources: Insee, Eurostat, the CCSS, the Banque de France, the World Inequality Database and the Institut des politiques publiques, each named in its panel.

The page is static HTML and reads in full without JavaScript. The script adds the tooltips, traces a flow through a diagram, opens in place the evidence kept on separate pages, shows and hides the diagrams on a phone, recomputes the wealth calculator, places you on the page that asks where you stand, and keeps the theme you pick: light, dark or high contrast.

Questions, ideas or corrections: open an issue at github.com/samoht/budget/issues. Who makes this site, and the corrections made to it, are on the Method page.

Every data file, with its columns, licence and checksum: the Data page. Every chart, by question: the index of the charts. Every line of the register: the register's index.

The evidence summaries of the register as they are written: Latest updates (Atom feed).

Text and data: Licence Ouverte / Open Licence 2.0, except third-party data, which keep their publishers' terms. Code: ISC. See LICENSE.md.

Version unreleased-ef7a7ba2, built on 2026-11-12 from ef7a7ba2.