Demographics and pensions: an ageing country

Pensions are one of the questions the parties will have to answer before 29 November. To understand the debate there are three starting figures: how much is spent, what share of earnings the pension replaces and for how many years it is drawn.

Spain spent 13.2% of GDP on pensions in 2024, against an EU average of 12.6%: the seventh highest share of the 27. In 2008 it was 9.2%, below the European average (11.8%). The rise, of 4.0 points, is the second largest in the EU after Finland's (4.2). The European definition includes old-age, survivors', orphans' and disability pensions, including those of civil servants.

One reason for that spending is the amount. The median pension of people aged 65 to 74 equals 81% of the median earnings of people aged 50 to 59, gross. In the EU, 60%. Only Luxembourg has a higher ratio (88%); Italy is just behind Spain (80%). It is a snapshot of two different groups in the same year, not the pension each worker will receive relative to their last wage, but it shows that retirement cuts income less in Spain than in almost all of Europe.

The other reason is duration. At 65, a person living in Spain can expect to live another 21.9 years on average, the same as in France and more than in any other EU country, where the average is 20.2. In 1975 it was 15.2 years: someone retiring at that age today draws their pension, on average, 6.7 years longer than then.

This lens across all themes

On 1 January 2025 there were 31.7 people aged 65 or over in Spain for every 100 aged 16 to 64. The figure ranges from 46.5 in Asturias, 44.7 in Castile and León and 43.7 in Galicia to 24.5 in the Balearic Islands, 25.7 in Murcia and 26.1 in the Canary Islands. In 1975, Asturias had 17.2, less than the Balearic Islands (19.5): since then it has added 29.3 points, the largest rise of the 17 regions, and the Balearic Islands 5.0, the smallest. Fertility, life expectancy and migration all play a part in that difference, both young people leaving and new residents arriving.

That ratio is demographic: it compares ages, not who works. Counting people in work of all ages for every person aged 65 or over, Spain had 2.19 in the last year, against 2.79 in 2008. The figure has fallen in all 17 regions. Since 2008 the number of people in work has grown 9.8% and the number over 64, 39.6%.

The regional differences remain: in the Balearic Islands there are 2.98 people in work for every person over 64 and in Asturias, 1.50. Pensions are paid by the common Social Security system, so these differences do not mean different pensions by region, but they do show where ageing weighs most on the services the regions run, such as healthcare and long-term care.

This lens across all themes

In 1975 Spain had 2.77 children per woman. In 1981 the indicator fell below 2.1 for the first time (2.04), the level at which a population stays stable without migration, and kept falling to 1.13 in 1998. It recovered to 1.44 in 2008 and has fallen again since: 1.10 in 2024, the lowest figure in the series. In the EU only Malta has fewer children per woman, and no country reaches 2.1.

Why is spending rising? Between 2008 and 2024, pension spending adjusted for inflation grew 53.5%. Two reasons add up: the number of pensioners rose 21.4%, to 10.4 million, and real spending per pensioner 26.4%. Roughly, 45% of the rise is because there are more pensioners and 55% because they receive more.

The average pension does not rise only because existing pensions are uprated. New retirees had, on average, higher wages and longer contribution records than the people whose pensions stop being paid. That is why the average pension can grow faster than inflation even though each pension is only uprated in line with it.

What comes next? According to INE's projections, the dependency ratio will go from 31.7 in 2025 to 51.8 in 2050. It would reach 52.2 in 2054, fall slightly and rise again to 53.1 in 2075.

INE's central scenario assumes that immigrants keep arriving at the pace of recent years. If as many people arrived as left, the ratio would reach 69.9 in 2050 and 90.4 in 2075. The difference between those two scenarios is much larger than between INE's own high and low fertility scenarios: of the assumptions INE tests, net migration is the one that moves the projection most.

See also: Immigration in data: who arrives, where they work and how much they weigh in the population

Policy file: the 2021 and 2023 pension reforms

1. What it is. Two laws. Law 21/2021 goes back to uprating pensions every year in line with average inflation. Royal Decree-law 2/2023 raises the system's revenue: an earmarked contribution, the Intergenerational Equity Mechanism (MEI), which starts at 0.6 points in 2023 and rises by a tenth a year to 1.2 in 2029; an increase in the maximum contribution base of 1.2 points a year above inflation between 2024 and 2050; an additional solidarity contribution on wages above that base, from 2025; and the option of calculating the pension over the last 29 years, discarding the 24 worst months.
2. Who and when. Parliament passed Law 21/2021 in December 2021. The Council of Ministers approved Royal Decree-law 2/2023 in March 2023 and Congress validated it.
3. What should happen. Pensions should keep their purchasing power and the new revenue should cover part of the rise in spending. The law includes a safeguard clause: every three years, AIReF (Spain's independent fiscal authority) checks whether average pension spending between 2022 and 2050, net of the new revenue, exceeds 13.3% of GDP. If it does, the government must propose measures or, failing that, raise the MEI.
4. What we compare it with. With the spending projections of AIReF and the European Commission, and with the rest of the public finances.

What the independent evaluations say. AIReF (Study 1/2026, May 2026) concludes that the rule is being met: spending net of the new revenue measures will average 13.0% of GDP between 2022 and 2050, below 13.3%. Those measures would bring in 1.6% of GDP a year on average, 0.4% from the MEI. But AIReF itself warns that the rule is not a good indicator of sustainability: it expects gross pension spending to go from 13% to 16.4% of GDP between 2025 and 2050, transfers from the State and other Social Security funds to pay for it to rise by 2.3 points of GDP, to 3.0%, and, if policies do not change, public debt to reach 123% of GDP in 2050.

Ángel de la Fuente (Fedea, April 2025) shares those doubts and questions how the rule was applied in its first evaluation: he criticises counting State transfers as system revenue and the upward revision of GDP forecasts, and points out that a tenth of a point either way would have changed the result. The evidence is therefore inconclusive: the rule is met, but there is no agreement that this guarantees the system is sustainable without new measures.

What this data does not say. Pension spending is measured in different ways: Eurostat's includes civil servants and is broader than Social Security's, and AIReF's does not match either, so the figures cannot be added to or subtracted from each other. People in work in the EPA are not the same as contributors registered with Social Security, nor does everyone over 64 draw a pension. The replacement ratio compares two different groups of people and does not say whether pensions are adequate, nor does it capture poverty among older people. Projections show what would happen if their assumptions held, not what will happen. And none of this data makes it possible to attribute the trend in spending to a particular government: it depends above all on demographics and on rules passed over decades.

This lens across all themes