Skip to content
Sverigefakta.com

Integration · data-driven

Integration over time

Does integration ’just take time’? We follow four arrival cohorts (1990–2016) and measure employment after 2, 5, 10 and 15 years in Sweden.

Sources: Statistics Sweden STATIV/Integration, Aldén & Hammarstedt (IFAU 2016), NIER 2024.

Last updated: View changelog

Employment rate per cohort

Each line is an arrival cohort. X-axis: years in Sweden. Y-axis: employment rate (ages 20–64, refugees and family-tied).

0%20%40%60%80%Swedish-born 20–64 yrs: ~83 %2 year5 year10 year15 year
1990–94 (Balkan)2000–042010–132014–16 (största kohorten)

Source: Statistics Sweden ‘Integration — labour-market establishment’ + Aldén & Hammarstedt (2016). Refugees + family-tied, ages 20–64.

Years in Sweden1990–942000–042010–132014–16
2 year22 %18 %14 %11 %
5 year48 %39 %32 %28 %
10 year62 %54 %47 %43 %
15 year67 %59 %53 %—

Reading: after 10 years in Sweden, the 1990–94 cohort was at 62 % employment. The 2014–16 cohort is at 43 %.

Welfare dependency by time in Sweden

Share (%) of ages 25–64 whose main source of income is public transfers (social assistance, establishment benefit, sickness/activity compensation, unemployment insurance).

Time in SwedenNon-European backgroundNordic backgroundNative-born
0–4 år76 %11 %7 %
5–9 år52 %9 %7 %
10–14 år38 %8 %7 %
15–19 år32 %8 %7 %
20+ år28 %8 %7 %

Source: Statistics Sweden STATIV 2022 (register data). Deep dive: welfare dependency over time

Three deep dives

FAQ

Does integration ’just take time’?
Later cohorts have lower employment after the same number of years in Sweden than the 1990s cohort. After 10 years the 2014–16 cohort sits at ~43 %, whereas the 1990–94 cohort was at ~62 %. Source: Statistics Sweden STATIV and Aldén & Hammarstedt (2016).
What counts as ‘employed’?
The LFS definition: at least 1 hour of paid work during the reference week, or absent from a job (sick leave, holiday, etc.). It is a broad measure including part-time work, subsidised jobs and self-employment.
Are the cohort differences ‘just business cycle’?
NIER (2024) concludes that educational structure, country of origin and arrival cycle jointly explain the pattern — but the gap remains even after adjustment. The 1990s Balkan cohort had higher average formal education, which is part of the explanation.
Why is the 15-year figure missing for the 2014–16 cohort?
The measurement window has not been reached — the cohort has been in Sweden for at most ~10 years as of the latest publication (2024).

Primary sources