Pensions: what the figures say

How much European countries spend on pensions, at what age people actually leave work, and the French trajectory projected by the COR — each figure with its definition and scope.

Scope: Eurostat and OECD European comparisons; the COR’s French trajectory (reference scenario), shown separately · Data as of Oct 30, 2026.

Pension expenditure, France15.0%% of GDP in 2024 (Eurostat) · EU: 12.6%
Exit age / normal age, France61.9 / 64.3men, 2024 (OECD)
Employment aged 55-64, France61.7%% in 2025 (Eurostat) · EU: 66.4%
People aged 20-64 per person 65+2.62France, 2023 (Insee, via the COR)

Key takeaways

  • On Eurostat’s harmonised definition, France spends 15.0 % of GDP on pensions in 2024, 3rd of the 27 EU countries behind Italy (15.9 %) and Austria (15.4 %), against 12.6 % for the EU (provisional). Ireland is at 4.1 %.
  • The French leave work early: 61.9 for men and 62.4 for women in 2024 (effective exit age, OECD), against a normal retirement age of 64.3 for a career started at 22.
  • The employment rate of people aged 55-64 reaches 61.7 % in 2025, up more than 23 points since 2003, but stays below the EU average (66.4 %); the gap is widest at 60-64 (44.4 % against 55.0 %).
  • The median income of people aged 65+ is 97 % of that of the under-65s in 2025 in France (91 % in the EU, 76 % in the Netherlands).
  • In the COR reference scenario there would be 1.62 people aged 20-64 per person aged 65+ in 2070, against 2.62 in 2023; pension spending would rise from 13.9 % of GDP in 2024 to 15.3 % in 2070 and retirees’ relative standard of living fall from 100.2 % (2023) to 90.3 %.

Briefing note: how to read these figures

The weight of pensions in the economy comes from three levers. Debates often mix them up; four pitfalls recur in almost every comparison.

The three levers

  1. How many retirees per worker. The demographic ratio (people aged 20-64 per person 65+) and the actual retirement age set how many pensions each worker funds. Raising the age of exit from work acts on both terms: fewer retirees, more contributors.
  2. The level of pensions. For a given number of retirees, spending depends on the average pension relative to earnings: that is what retirees’ relative standard of living measures, and what indexation and benefit rules act on.
  3. The public share. Part of pensions can go through mandatory or voluntary private funds (Netherlands, Sweden): public spending is then lower, without the total effort being smaller.

The four pitfalls

  1. Public spending ≠ total spending. According to the OECD, the Netherlands spends 7.5 points of GDP on public old-age and survivors’ benefits but 3.1 points on private ones (2021); France 13.8 points public and 0.2 private (2022). Comparing public spending alone compares different systems.
  2. Legal age ≠ exit age ≠ claiming age. The legal (or « normal ») age is a rule; the labour-market exit age (OECD) is what employment shows; the retirement age (COR) is the age at the first pension. They differ by several years and are not substitutes.
  3. Eurostat and the COR have different scopes. Eurostat counts old-age, survivors’, disability and early-retirement pensions, public and private; the COR only the mandatory French schemes (incl. FSV, excl. RAFP). For 2024 Eurostat gives 15.0 % of GDP and the COR 13.9 %: not a contradiction but two different measures, shown here on separate charts.
  4. A projection is a scenario. The COR curves beyond 2025 depend on assumptions (productivity, unemployment, net migration, fertility). They describe what would happen if those assumptions held, not what will happen: they are dashed and never shown as the headline value.

Pension expenditure as % of GDP, EU countries

Eurostat scope: old-age, survivors’, disability and early-retirement pensions; public and private spending under the Eurostat definition (ESSPROS). Ranking for the selected year. Comparable across European countries; not to be set against the COR figures (French mandatory schemes only).

Year

* provisional or estimated · Eurostat spr_exp_pens

Private share of pensions (OECD)

Old-age and survivors’ benefits, % of GDP: public and private (mandatory and voluntary). Last year published by the OECD (2021-2022) — a source and definition distinct from Eurostat’s, on a separate chart.

  • Public
  • Private
  • Netherlands (2021)Public 7.5% · Private 3.1%
  • Sweden (2021)Public 10.3% · Private 2.0%
  • Italy (2021)Public 16.2% · Private 0.8%
  • Germany (2021)Public 11.6% · Private 0.7%
  • Spain (2021)Public 12.7% · Private 0.3%
  • France (2022)Public 13.8% · Private 0.2%

OCDE / OECD — SOCX (DSD_SOCX_AGG@DF_PUB_PRV, TP01)

Effective labour-market exit age and normal retirement age

OECD, Pensions at a Glance, 2024. Filled dot: average effective age of exit from the labour market. Circle: current normal retirement age for a full career started at 22. Neither is the age at which the pension is claimed.

Sex
  • Effective exit age
  • Normal retirement age
6061626364656667Estonia — Effective exit age 66.9 · Normal retirement age 64.8Estonia66.9 · 64.8Portugal — Effective exit age 66.5 · Normal retirement age 65.3Portugal66.5 · 65.3Denmark — Effective exit age 66.3 · Normal retirement age 67.0Denmark66.3 · 67.0Ireland — Effective exit age 66.1 · Normal retirement age 66.0Ireland66.1 · 66.0Cyprus — Effective exit age 65.6 · Normal retirement age 65.0Cyprus65.6 · 65.0Sweden — Effective exit age 65.1 · Normal retirement age 66.0Sweden65.1 · 66.0Netherlands — Effective exit age 65.0 · Normal retirement age 67.0Netherlands65.0 · 67.0Bulgaria — Effective exit age 64.8 · Normal retirement age 64.6Bulgaria64.8 · 64.6Czechia — Effective exit age 64.2 · Normal retirement age 64.2Czechia64.2 · 64.2Germany — Effective exit age 64.2 · Normal retirement age 66.2Germany64.2 · 66.2Italy — Effective exit age 64.0 · Normal retirement age 64.8Italy64.0 · 64.8EU-27 — Effective exit age 63.9 · Normal retirement age 64.7EU-2763.9 · 64.7Greece — Effective exit age 63.7 · Normal retirement age 62.0Greece63.7 · 62.0Poland — Effective exit age 63.7 · Normal retirement age 65.0Poland63.7 · 65.0Hungary — Effective exit age 63.6 · Normal retirement age 65.0Hungary63.6 · 65.0Latvia — Effective exit age 63.6 · Normal retirement age 64.8Latvia63.6 · 64.8Finland — Effective exit age 63.5 · Normal retirement age 65.0Finland63.5 · 65.0Lithuania — Effective exit age 63.5 · Normal retirement age 64.7Lithuania63.5 · 64.7Romania — Effective exit age 63.5 · Normal retirement age 65.0Romania63.5 · 65.0Austria — Effective exit age 62.7 · Normal retirement age 65.0Austria62.7 · 65.0Slovakia — Effective exit age 62.7 · Normal retirement age 63.2Slovakia62.7 · 63.2Croatia — Effective exit age 62.6 · Normal retirement age 65.0Croatia62.6 · 65.0Malta — Effective exit age 62.5 · Normal retirement age 64.0Malta62.5 · 64.0Belgium — Effective exit age 62.4 · Normal retirement age 65.0Belgium62.4 · 65.0Spain — Effective exit age 62.4 · Normal retirement age 65.0Spain62.4 · 65.0France — Effective exit age 61.9 · Normal retirement age 64.3France61.9 · 64.3Slovenia — Effective exit age 60.8 · Normal retirement age 62.0Slovenia60.8 · 62.0Luxembourg — Effective exit age 60.1 · Normal retirement age 62.0Luxembourg60.1 · 62.0

OCDE / OECD — Pensions at a Glance (DSD_PAG@DF_PAG)

Employment rate of older workers

Share of people in employment, Labour Force Survey (Eurostat). 55-64: lfsi_emp_a; 60-64: lfsa_ergan.

Country
  • Aged 55-64
  • Aged 60-64

France, Spain: definition differs (Eurostat flag); France: break in series in 2024.

Relative median income of people aged 65+

Median equivalised disposable income of people aged 65+ divided by that of the under-65s (1 = parity). EU-SILC, survey year (incomes of the previous year). A different measure from the COR’s mean standard of living of retirees.

  • France
  • Germany
  • Italy
  • Spain
  • Sweden
  • Netherlands
  • EU-27

1 = parity · Eurostat ilc_pnp2

The French trajectory according to the COR

A different source: the French pensions advisory council (COR), mandatory French schemes. These figures are comparable neither with the Eurostat and OECD charts above nor with other countries.

COR reference scenario (annual report, June 2026), sensitive to the productivity, unemployment and net-migration assumptions: fertility 1.45 from 2028, net migration +150,000 a year, hourly productivity +0.7 % a year, unemployment 7 % from 2040.

  • Observed
  • Estimate / Projection

Demographic ratio (aged 20-64 / 65+)

2.62last observed year: 2023

Insee; 2024-2026 provisional (dashed), then the central demographic projection. Projection 2070 — 1.62.

Pension spending, % of GDP

13.9%last observed year: 2024

Mandatory schemes, incl. FSV, excl. RAFP. 2025: CCSS estimate (dashed). Projection 2070 — 15.3%.

Average age at retirement

63.0 yrslast observed year: 2024

Period age at first pension (Drees). Same projection in every scenario. Projection 2070 — 64.6 yrs.

Retirees’ standard of living, % of the whole population

100.2%last observed year: 2023

Mean standard of living relative to the whole population’s (100 = parity). Sparse observations before 1996. Projection 2070 — 90.3%.

Pension system balance, % of GDP

−0.1%last observed year: 2024

Resources minus spending, EPR convention. 2025: estimate (dashed). Projection 2070 — −2.4%.

Data

Show pension expenditure by country (Eurostat)
Eurostat scope: old-age, survivors’, disability and early-retirement pensions; public and private spending under the Eurostat definition (ESSPROS).
Country201420192024
Italy16.4%15.9%15.9%
Austria14.9%14.1%15.4%
France15.2%14.9%15.0%
Finland13.4%13.4%14.5%
Greece17.4%15.8%14.1%
Belgium12.3%12.6%13.5%
Spain12.7%12.6%13.2%
Portugal15.6%13.7%12.9%
EU-2713.1%*12.6%*12.6%*
Germany11.5%11.7%11.7%*
Poland11.8%10.8%11.6%
Netherlands12.0%11.3%11.3%
Denmark14.0%12.7%11.2%
Sweden11.6%10.8%10.7%
Luxembourg9.2%9.7%10.3%
Slovenia11.3%9.7%10.0%
Bulgaria8.7%7.5%9.6%
Slovakia8.6%8.3%9.5%
Croatia10.7%*9.9%9.4%
Romania8.2%*7.8%9.1%
Czechia8.8%8.2%9.0%
Latvia8.2%*7.8%*8.9%*
Estonia7.4%7.6%8.7%
Cyprus10.0%8.7%7.6%
Lithuania7.0%*6.9%*7.3%
Hungary8.9%7.2%7.0%
Malta7.8%6.0%5.1%
Ireland7.4%5.0%4.1%

* provisional or estimated

Show the COR series (observed, estimate, projection)
A different source: the French pensions advisory council (COR), mandatory French schemes. These figures are comparable neither with the Eurostat and OECD charts above nor with other countries.
YearDemographic ratio (aged 20-64 / 65+)Pension spending, % of GDPAverage age at retirementRetirees’ standard of living, % of the whole populationPension system balance, % of GDP
20003.69—60.7102.4%—
20013.67—60.8101.4%—
20023.6411.8%60.8101.3%0.3%
20033.6211.9%60.8101.7%0.4%
20043.6112.0%60.6101.6%0.4%
20053.5812.1%60.6102.8%0.1%
20063.5812.2%60.5103.0%0.1%
20073.5912.3%60.5102.0%0.0%
20083.5812.4%60.4101.8%−0.1%
20093.5613.3%60.5102.1%−0.5%
20103.5313.4%60.5102.2%−0.8%
20113.5013.5%60.7102.6%−0.7%
20123.4013.8%61.0104.3%−0.6%
20133.3014.0%61.2104.7%−0.4%
20143.2014.1%61.4103.9%−0.3%
20153.0914.0%61.7103.9%−0.3%
20163.0014.0%61.9103.4%−0.2%
20172.9213.9%62.1101.3%−0.0%
20182.8513.8%62.299.9%−0.1%
20192.7913.7%62.399.9%−0.0%
20202.7314.7%62.499.4%−0.6%
20212.6913.8%62.698.7%−0.0%
20222.6613.6%62.797.7%0.1%
20232.6213.5%62.7100.2%0.0%
20242.58 (estimate)13.9%63.0102.3% (projection)−0.1%
20252.53 (estimate)14.1% (estimate)63.1 (estimate)103.2% (projection)−0.2% (estimate)
20262.49 (estimate)14.1% (projection)63.3 (projection)103.2% (projection)−0.2% (projection)
20272.45 (projection)14.2% (projection)63.4 (projection)102.9% (projection)−0.2% (projection)
20282.41 (projection)14.2% (projection)63.6 (projection)102.7% (projection)−0.2% (projection)
20292.37 (projection)14.1% (projection)63.8 (projection)102.3% (projection)−0.2% (projection)
20302.33 (projection)14.1% (projection)63.9 (projection)102.2% (projection)−0.2% (projection)
20312.29 (projection)14.0% (projection)64.1 (projection)101.3% (projection)−0.2% (projection)
20322.26 (projection)14.0% (projection)64.2 (projection)100.6% (projection)−0.3% (projection)
20332.23 (projection)14.0% (projection)64.3 (projection)99.5% (projection)−0.3% (projection)
20342.20 (projection)14.0% (projection)64.3 (projection)98.7% (projection)−0.3% (projection)
20352.18 (projection)14.0% (projection)64.3 (projection)98.0% (projection)−0.4% (projection)
20362.15 (projection)14.0% (projection)64.4 (projection)97.5% (projection)−0.4% (projection)
20372.12 (projection)14.0% (projection)64.4 (projection)97.3% (projection)−0.5% (projection)
20382.10 (projection)14.0% (projection)64.5 (projection)97.2% (projection)−0.5% (projection)
20392.07 (projection)14.0% (projection)64.5 (projection)97.1% (projection)−0.6% (projection)
20402.06 (projection)14.0% (projection)64.5 (projection)96.9% (projection)−0.6% (projection)
20412.05 (projection)14.1% (projection)64.5 (projection)96.8% (projection)−0.7% (projection)
20422.04 (projection)14.1% (projection)64.5 (projection)96.6% (projection)−0.7% (projection)
20432.04 (projection)14.1% (projection)64.5 (projection)96.5% (projection)−0.8% (projection)
20442.03 (projection)14.1% (projection)64.5 (projection)96.4% (projection)−0.8% (projection)
20452.01 (projection)14.2% (projection)64.5 (projection)96.3% (projection)−0.9% (projection)
20461.99 (projection)14.2% (projection)64.5 (projection)96.0% (projection)−1.0% (projection)
20471.98 (projection)14.2% (projection)64.5 (projection)95.7% (projection)−1.0% (projection)
20481.96 (projection)14.3% (projection)64.5 (projection)95.5% (projection)−1.1% (projection)
20491.94 (projection)14.3% (projection)64.5 (projection)95.3% (projection)−1.1% (projection)
20501.93 (projection)14.4% (projection)64.5 (projection)95.1% (projection)−1.2% (projection)
20511.91 (projection)14.4% (projection)64.6 (projection)95.0% (projection)−1.3% (projection)
20521.89 (projection)14.5% (projection)64.6 (projection)94.8% (projection)−1.4% (projection)
20531.87 (projection)14.5% (projection)64.6 (projection)94.6% (projection)−1.4% (projection)
20541.86 (projection)14.6% (projection)64.6 (projection)94.5% (projection)−1.5% (projection)
20551.84 (projection)14.6% (projection)64.6 (projection)94.3% (projection)−1.5% (projection)
20561.82 (projection)14.6% (projection)64.6 (projection)94.1% (projection)−1.6% (projection)
20571.81 (projection)14.7% (projection)64.6 (projection)93.9% (projection)−1.6% (projection)
20581.79 (projection)14.7% (projection)64.6 (projection)93.7% (projection)−1.7% (projection)
20591.78 (projection)14.7% (projection)64.6 (projection)93.6% (projection)−1.7% (projection)
20601.77 (projection)14.7% (projection)64.6 (projection)93.2% (projection)−1.8% (projection)
20611.76 (projection)14.8% (projection)64.6 (projection)93.0% (projection)−1.8% (projection)
20621.74 (projection)14.8% (projection)64.6 (projection)92.8% (projection)−1.9% (projection)
20631.73 (projection)14.9% (projection)64.6 (projection)92.6% (projection)−1.9% (projection)
20641.71 (projection)15.0% (projection)64.6 (projection)92.3% (projection)−2.0% (projection)
20651.70 (projection)15.0% (projection)64.6 (projection)92.0% (projection)−2.1% (projection)
20661.68 (projection)15.1% (projection)64.6 (projection)91.7% (projection)−2.2% (projection)
20671.66 (projection)15.2% (projection)64.6 (projection)91.4% (projection)−2.3% (projection)
20681.64 (projection)15.2% (projection)64.6 (projection)91.0% (projection)−2.3% (projection)
20691.63 (projection)15.3% (projection)64.6 (projection)90.6% (projection)−2.4% (projection)
20701.62 (projection)15.3% (projection)64.6 (projection)90.3% (projection)−2.4% (projection)

Sources and method

  • Eurostat, pension expenditure. European system of integrated social protection statistics (ESSPROS), dataset spr_exp_pens, all pension types, means-tested or not, % of GDP. « p » = provisional, « e » = estimated.Eurostat — spr_exp_pens
  • Eurostat, employment and income. Labour Force Survey (employment rates aged 55-64 and 60-64) and EU-SILC (relative median income of people aged 65+).Eurostat — lfsi_emp_aEurostat — lfsa_erganEurostat — ilc_pnp2
  • OECD. Pensions at a Glance (average effective age of labour-market exit; current normal retirement age, career started at 22) and the SOCX database (public and private old-age and survivors’ benefits).OECD Data Explorer — Pensions at a GlanceOECD Data Explorer — SOCX
  • COR. Annual report « Évolutions et perspectives des retraites en France », June 2026, data files (synthesis and part 1), reference scenario only. Licence Ouverte / Etalab 2.0. The 2025 point of spending, balance and age is an estimate; Insee’s 2024-2026 population estimates are provisional.COR — rapport annuel juin 2026
  • Guard-rails. Every series is checked when the snapshot is built: unit, plausibility bounds, internal coherence (balance = resources − spending; EU aggregate within its members’ range; employment 60-64 ≤ 55-64) and concordance with the published values. A series out of bounds is dropped, never corrected.