OWN THE FUTURE.

OWN THE FUTURE · SEASON 2 · DEMOGRAPHICS & SOCIAL CHANGE

S2:E4Ageing

How many of us are old at the same time? Counted from 1900 to 2025

In many countries the share of older people is larger than it was a generation ago, and it shows in everything from housing to the labour market. We explain what the figures show and which words demographers use.

Published 10 Sep 2026 · About 9 minutes to read

In 1913 the Swedish Parliament decided on Sweden's first general pension insurance.10, 14 It was compulsory, and the pension age was set at 67.14 In 2026 the target retirement age for the public pension is 67.9 The same figure, 113 years apart.

The population behind the figure is a different one. At the end of 1913, 469,625 people in Sweden were 65 or older, so 8.33%.1 At the end of 2025 there were about 2,235,500, so 21.08%.1, 4 The share is two and a half times as large.1

This article counts that change in Swedish statistics and goes through the words demographers use when they do. How long an individual person lives belongs to S2:E2 Longevity. This is about something else: how many of us are old at the same time.

What shapes the age structure

Age structure is the demographers' term for how a population is made up by age. It is decided first of all by how many people were born in different years, and only after that by how long each of them lives. A large group of children born in one year becomes a large group of pensioners sixty-five years later, and a small group becomes a small group.5 Migration moves the figures, but the size of each birth cohort gives the curve its shape.

Sweden's number of births swung sharply in the first half of the 20th century. In 1920, 138,753 children were born.3 Then the number fell, year after year with a single exception, to 85,020 in 1933, the lowest figure in the whole of the 20th century.3 After that it turned upwards again and reached 135,373 in 1945.3 That movement shows up in the age structure half a century later.

The curve is not smooth

In 1900, 8.37% of the population was 65 or older.1 In 1950 the share was 10.24%, in 1970 it was 13.78 and in 1980 it was 16.38.1 Up to then the trend looks like an even slope upwards.

Then it levelled off and turned down. The share reached 17.80% in 1989 and fell to 17.16% in 2002, lower than it had been thirteen years earlier.1 Only after that did it start to rise again, every year, to 21.08% in 2025.1

The explanation lies in the number of births. Those who turned 65 between 1989 and 2002 were born between 1924 and 1937, during the fall towards the low point of 1933.3 In those years unusually few people joined the group aged 65 and older. Those born during the rise towards 1945 turned 65 around 2010 and are turning 80 now.3 That is why the oldest group is growing during the 2020s in particular.

Share of the population, per cent65 and older80 and older8.371.07190010.241.52195013.782.35197017.804.21198917.165.25200221.086.452025The plateau 1989 to 2002: the share aged 65 and older fell over thirteen years.
The share of Sweden's population aged 65 and older and aged 80 and older, six selected years. The editors' own sum of Statistics Sweden's single-year age groups by age and sex. The share aged 65 and older rose from 8.37% in 1900 to 17.80% in 1989, fell to 17.16% in 2002 and then rose to 21.08% in 2025. The share aged 80 and older went from 1.07 to 6.45% over the same period. Schematic figure. Source: Statistics Sweden, Statistical Database, population by age and sex, the tables BefolkningR1860N for 1860 to 2024 and BefolkningR1860NCKM for 2025, retrieved 9 Sep 2026.

Two thresholds were passed recently. The number of people aged 65 and older passed two million in 2017, when there were 2,006,146.1 The share passed 20% in 2020, when it was 20.12% against 20.00 the year before.1

The dependency ratio, and what it counts

The ratio today, and how it is counted

The measure most often used to describe what the age structure means is called the demographic dependency ratio.15 Statistics Sweden counts it as the number of people aged 0 to 19 plus the number aged 65 and older, per 100 people aged 20 to 64.15, 2 The measure says nothing about who works.15 It counts ages.

For 2025 the ratio is 77.0.2 Of that, 37.3 comes from people aged 65 and older and 39.7 from people under 20. In 2000 the same figure was 70.4 in total, 29.4 from older people and 41.0 from younger people.2

There is a difference here that often disappears in summaries. The older part rose by 7.9 units over the 25 years.2 The total rose by 6.6, because the younger part fell by 1.3 at the same time.2

The same ratio counted over a century

Counted over a century, the figure goes the other way. In 1900 the total dependency ratio was 101.0, so there were more people outside ages 20 to 64 than inside them.1 Of that, 16.8 came from older people and 84.2 from children.1 In 1970 the total was 70.7.1 Sweden has gained more older people and fewer children at the same time, and the total has fallen.

A figure that is easier to remember: in 1970 there were 4.25 people aged 20 to 64 for every person aged 65 and older.1 In 2000 the figure was 3.40 and in 2025 it is 2.68.1 Counted against people aged 80 and older, there were 24.9 in 1970 and 8.8 in 2025.1

21.08%

of the population in Sweden was 65 or older on 31 December 2025. A generation ago, in 1995, the share was 17.46%.

SOURCE: STATISTICS SWEDEN, STATISTICAL DATABASE, OWN SUM

What the ratio does not capture

The measure counts everyone over 64 as supported by others, and that fits less well every year. The employment rate among 65 to 74-year-olds was 9.7% in 2001 and 22.2% in 2025 according to the labour force surveys, and the number in work at those ages rose from 71,900 to 231,900.7 The survey was redesigned in 2021 to fit an EU regulation, which creates a break in the series.16, 7 So the comparison between 2001 and 2025 should be read with that in mind.16 Among 55 to 64-year-olds the employment rate went from 66.2 to 78.3%.7

The OECD measures the same thing from the other direction, as the actual age of leaving the labour market. For Sweden it was 65.1 years for men and 65.0 years for women in 2024, against 62.4 and 61.5 years respectively in 2000.13 The OECD average in 2024 was 64.7 and 63.6 years.13

The oldest are growing fastest

The same birth cohorts move through the higher ages fifteen years later, so the pattern repeats itself. The share aged 80 and older reached 5.38% in 2005, fell to 5.07% in 2017 and then rose to 6.45% in 2025.1 The small cohorts from the 1930s turned 80 in the years around 2015, and the large cohorts from the 1940s are turning 80 now.

The number in the group has grown over the longer period, even in the years when the share fell, but it has dropped in single years. The group was 498,218 people in 2011, 498,148 in 2012 and 497,717 in 2013.1 In 2015, 501,650 people were 80 or older.1 In 2025 there were about 684,000, so 36.3% more in ten years.1

At the very top of the age range the change is larger than anywhere else. In 1960 there were 9,621 people in Sweden aged 90 or older, 0.13% of the population.1 In 2025 there were about 107,300, so 1.01%.1 In 1970 there were 127 people aged 100 or older.1 In 2025 there were about 3,000.

Portrait of a very old man in a suit, waistcoat and tie, sitting by a window in a room with patterned wallpaper, black and white
Nilsson, a bricklayer from Långlöt, celebrates his 100th birthday at the Halltorp old people's home. The museum record is undated. In 1970 there were 127 people in Sweden aged 100 or older. In 2025 there were about 3,000. Photograph: Allan Bernving. Kalmar County Museum, KLMF.Bernvingålderdom8. Public domain.

Care counted in outcomes

What care for older people costs

Every year the National Board of Health and Welfare (Socialstyrelsen) reports what health and social care for older people costs and covers.8 For 2024 the municipalities' costs came to 166 billion kronor, equal to 2.6% of GDP.8 Special housing accounted for 57% of the cost, and care for older people for about 52% of the municipalities' total costs for social services.8

Between 2020 and 2024 costs rose by 5% in fixed prices, while the number of people with at least one service rose by 7.6%.8 The cost per person receiving services was about 475,000 kronor in 2024 and has fallen slightly since 2020.8

Who gets care, and how late

The share of the oldest who receive care has fallen. Between 2015 and 2025 the share of people aged 80 and older with home help went from just over 20 to about 16%, and the share living in special housing from just under 13 to about 9%.8 At the same time the group aged 80 and older grew by 36.3%, from 501,650 people in 2015 to about 684,000 in 2025, so the falling shares do not mean that the number fell.1

Care is given late. The median age on moving into special housing is just over 86, 87.1 years for women and 84.2 for men, and the median length of stay is 27 months.8 The share who had home help before they moved in rose from 70% in 2015 to 79% in 2025, so from about 17,300 to 25,700 people.8 In 2024, 114,350 assistant nurses worked in care for older people, 75,810 of them in special housing, together with 73,800 care assistants.8

Sweden compared with other countries

Sweden aged early. As early as 1970, 13.78% of the population was 65 or older.1 Europe as a whole reached that level only in the mid-1990s, a quarter of a century later.12 According to the UN's population estimates, the share in Europe was 7.89% in 1950, 14.71% in 2000 and 20.09% in 2023, and for the world as a whole 5.07, 6.83 and 10.00% in the same years.12

Today Sweden is below the European average. Eurostat gives the share aged 65 and older as of 1 January 2025 as 20.8% in Sweden, 22.0% in the EU27, 22.7% in Germany, 24.7% in Italy and 15.7% in Ireland.11 The median age was 41.2 years in Sweden against 44.9 years in the EU27 and 49.1 years in Italy.11 Eurostat's reference date of 1 January 2025 is the same population that Statistics Sweden reports as of 31 December 2024, which is why the figure is given as 20.8 and not 21.08.11, 1

The OECD measures the ratio between people aged 65 and older and people aged 20 to 64, the same definition Statistics Sweden uses.13, 15 For Sweden the figure was 36.8 in 2024, against 32.6 in the average OECD country and 34.4 in the EU27.13 In 1980 Sweden was at 28.4.13

What the agencies project

Everything above has been measured. What concerns the future is someone else's projection and should be read as such. In its population projection published on 16 April 2026, Statistics Sweden says that the number of people aged 85 and older rises from 304,000 in 2025 to 483,000 in 2035, that the number aged 6 to 15 falls from 1,238,000 to 1,055,000 and that the population grows by 170,000 to 10.8 million.5

Statistics Sweden itself explains what such a figure is. In the short term, the agency writes, the projection can be seen as a population forecast, and in the longer term it should be seen as exactly that: a projection of the population if the behaviour and trends we can see today continue into the future.5

Statistics Sweden also shows what the calculation depends on. Its projection of 16 April 2026 assumes that all Ukrainian citizens with temporary protection emigrate in 2027, and it writes at the same time that the Temporary Protection Directive has been extended before and that it is very possible this will happen again.5, 17 In the same projection of 16 April 2026, Statistics Sweden gives two alternative calculations, which give just under 25,000 and just over 35,000 more inhabitants respectively in five to ten years.5, 17

A figure about the future does not have to be a projection. The target retirement age for the public pension is 67 for people born from 1960 to 1965 and applies from 2026 to 2032.9 It is set six years before it takes effect and is linked to how long we live.9, 18 It is a decided rule, not a calculation.18

What the figures do not say

Share and number are two different things. Between 1995 and 2025 the number of people aged 65 and older rose by 44.9%, from 1,543,332 to about 2,235,500.1 The share rose by 3.62 percentage points, from 17.46 to 21.08%.1 Both figures are correct, and the difference is because the population grew by 20.0% at the same time.1

Reference dates differ between the sources. Statistics Sweden reports as of 31 December, Eurostat as of 1 January the following year and the UN as of the middle of the year.1, 11, 12 So three sources can give three different figures for what looks like the same thing. Since 2025 Statistics Sweden has also protected population statistics with a method the agency itself calls the Cell Key Method.6 Statistics Sweden describes it like this: counts greater than zero are adjusted by a small negative or positive whole number, or left unchanged, this applies to all totals as well, and the adjustment is chosen at random from the numbers minus three to plus three.6 So exact numbers of people for 2025 are given rounded in this article.

And the dependency ratio describes ages, not support. It says nothing about who works, who pays tax or who needs help.15 It is a rough measure that is useful precisely because it is rough, and it should be read as a way of describing how a population is made up.

What to take away

At the end of 2025, 21.08% of the population in Sweden was 65 or older, against 8.37% in 1900, and the share aged 80 and older was 6.45%.1 The demographic dependency ratio was 77.0 for 2025, of which 37.3 came from people aged 65 and older.2 The force is slow and structural: the age make-up is decided by how large the cohorts already born are, and the large cohorts from the 1940s are turning 80 now.3 Follow Statistics Sweden's population by age and the demographic dependency ratio, both updated every year.1, 2 Also follow the National Board of Health and Welfare's annual status report on health and social care for older people, and the employment rate among 65 to 74-year-olds in Statistics Sweden's labour force surveys.7, 8

The next episode is S2:E5 Digital Health, on remote care and how the visits are counted.

Education, not advice.

The fund

Behind the library stands a fund.

Own The Future is published by AIX. The library is the knowledge; AIX Dynamic is the tool, a systematic securities fund (UCITS) that follows the trend and steps aside when risk rises.

This is marketing. Fund units can both rise and fall in value. Read the key information document (KID) and the prospectus before you invest; you will find them on the fund's page at Avanza.

34 UNITHOLDERS · AS OF 15 SEP 2026

Sources

  1. Statistics Sweden (Statistiska centralbyrån, SCB), Statistical Database (Statistikdatabasen), Folkmängden efter ålder och kön (Population by age and sex), the tables BefolkningR1860N for 1860 to 2024 (updated 21 Feb 2025) and BefolkningR1860NCKM for 2025 (updated 24 Feb 2026). All shares and numbers per age group, and the dependency ratios calculated backwards before Statistics Sweden's published series.
  2. Statistics Sweden, Statistikdatabasen (Statistical Database), Demografisk försörjningskvot efter region, år 2000 till 2025 (Demographic dependency ratio by region, 2000 to 2025) (FkvotHVD), updated 24 Mar 2026. The ratios 77.0, 37.3 and 39.7 for 2025 and 70.4, 29.4 and 41.0 for 2000.
  3. Statistics Sweden, Statistikdatabasen (Statistical Database), Befolkningsutvecklingen i riket efter kön, år 1749 till 2025 (Population change in the country by sex, 1749 to 2025) (BefUtvKon1749). The number of live births in 1920, 1933 and 1945.
  4. Statistics Sweden, Befolkningen växer trots fortsatt lågt barnafödande 2025 (The population grows despite continued low birth rates in 2025), statistical news. The population at the end of 2025.
  5. Statistics Sweden, Antalet födda fortsätter att minska (The number of births continues to fall), statistical news on Sweden's future population 2026 to 2070, published 16 Apr 2026. The projection for the group aged 85 and older, the agency's own description of what a projection is, and the assumption about temporary protection.
  6. Statistics Sweden, Frågor och svar om röjandeskyddet Cell Key Method (CKM) (Questions and answers on the Cell Key Method (CKM) disclosure protection). The method, that the adjustment is chosen at random from the numbers minus three to plus three, and that it applies to totals as well.
  7. Statistics Sweden, Statistikdatabasen (Statistical Database), Befolkningen 15 till 74 år (AKU) efter arbetskraftstillhörighet, kön och ålder, år 2001 till 2025 (Population aged 15 to 74 (Labour Force Survey) by labour force status, sex and age, 2001 to 2025) (AKURLBefAr), updated 23 Jan 2026. The employment rate among 65 to 74-year-olds and 55 to 64-year-olds. The survey was redesigned in 2021.
  8. National Board of Health and Welfare (Socialstyrelsen), Vård och omsorg för äldre, lägesrapport 2026 (Health and social care for older people, status report 2026), article number 2026-3-10088, published March 2026. The costs, the shares with home help and special housing, the median age on moving in and the staff numbers.
  9. Swedish Pensions Agency (Pensionsmyndigheten), Pensionsåldrar och riktålder (Pension ages and target retirement age). The target retirement age of 67 for people born from 1960 to 1965, and that it is set six years in advance.
  10. Swedish Pensions Agency, Pensionsåldern (The pension age), report 2011, chapter 2. The parliamentary decision of 1913 on a compulsory general pension insurance with a pension age of 67.
  11. Eurostat, Population structure indicators (demo_pjanind), the indicators PC_Y65_MAX and MEDAGEPOP, and the metadata for population statistics for the reference date. The shares and median ages as of 1 Jan 2025.
  12. United Nations, Department of Economic and Social Affairs, Population Division, World Population Prospects 2024, the medium variant. The share aged 65 and older in the world and in Europe in 1950, 2000 and 2023. The estimates of actual outcomes run up to and including 2023.
  13. OECD, Pensions at a Glance 2025, the dataflow DSD_PAG. The ratio between people aged 65 and older and people aged 20 to 64, and the actual age of leaving the labour market.
  14. The Swedish Parliament (Sveriges riksdag), Riksdagens skrivelse Nr 128 i anledning av Kungl. Maj:ts proposition med förslag till lag om allmän pensionsförsäkring (Parliament's communication no. 128 on the government's bill proposing an act on general pension insurance), 28 May 1913. Supports the parliamentary decision of 1913, that the insurance covered every Swedish man or woman, and that the right to a pension began at the latest on turning sixty-seven.
  15. Statistics Sweden, Statistical Database, Demografisk försörjningskvot efter region, år 2000 till 2025 (FkvotHVD) (Demographic dependency ratio by region, 2000 to 2025), updated 24 Mar 2026. Supports the definition of the measure, that the working-age population in the table means ages 20 to 64, that the measure is purely demographic and does not take into account how many people actually work, and the partial ratios from older and from younger people.
  16. Statistics Sweden, Förändringar i AKU (Changes in the Labour Force Survey), in-depth information on the labour force surveys. Supports that on 1 Jan 2021 the Labour Force Survey was adapted to the EU framework regulation, regulation 2019/1700, and that the adaptation caused breaks in the survey's time series.
  17. Statistics Sweden, Sveriges framtida befolkning 2026 till 2070 (Sweden's future population 2026 to 2070), Demografiska rapporter 2026:3, published 16 Apr 2026 (PDF). Supports the assumption that everyone with a residence permit under the Temporary Protection Directive emigrates in 2027, the agency's own wording that the directive has been extended before and that it is very possible this will happen again, and the two supplementary calculations of just under 25,000 and just over 35,000 more inhabitants respectively in five to ten years.
  18. The Swedish Parliament, Socialförsäkringsbalk (2010:110), 2 kap. 10 a till 10 d §§ om riktålder för pension (Social Insurance Code, Chapter 2, Sections 10 a to 10 d on the target retirement age). Supports that the target retirement age is calculated from the base figure 65 and the remaining life expectancy at 65, and that the calculated target age applies to the sixth year after the year of calculation, so that it is a decided rule and not a projection.

Links retrieved 9 Sep 2026.

Extended on 13 Sep 2026 with the parliamentary decision of 1913, the Statistical Database table behind the demographic dependency ratio, Statistics Sweden's description of the redesign of the labour force surveys in 2021, the population projection report and the legal provision on the target retirement age; the Swedish Pensions Agency says that the target retirement age of 67 applies from 2026 to 2032, while the article writes 2031, and Statistics Sweden's statistical news gives the change at ages 6 to 15 as minus 182,000, while the levels reported differ by 183,000.