Loading school data…
The link between income and school results
A municipality's socioeconomic profile co-varies strongly with school results — what that means and does not mean.
Get the main pattern quickly, find a local angle and jump straight to the charts that carry the story.
Statistics story · 3 visualisations · copy reviewed 18 October 2026 · Skolverket, Kolada and SCB.
One of the clearest patterns in Swedish school statistics is how strongly school results co-vary with the socioeconomic profile of a municipality's residents. Three visualisations show the correlation at municipal level, and why it should not be read as a claim about individual pupils or schools.
Socioeconomic background and merit score
Municipalities with high median income and high education levels usually have a higher average merit score (r ≈ 0.65 and 0.70 for 289 municipalities). Of the 52 municipalities in the top quarter on both income and education, 47 have a merit score above the median. The chart opens on the share of residents with a foreign background; choose median income or post-secondary education in the menu to see those correlations. The colour shows that the independent-school share is higher where income is higher (r ≈ 0.4).
Each bubble a municipality. Size: number of schools. Colour: independent school share. Choose the measure in the menu.
Open Socioeconomic correlation fullscreenEducation level — where do people with post-secondary education live?
The map shows the share of 25–64-year-olds with post-secondary education (2025). Danderyd, Lund, Lomma, Solna and several Stockholm suburbs are highest at 70–80 per cent. Filipstad is lowest at 24 per cent.
Share of 25–64-year-olds with post-secondary education per municipality — the darker, the higher the share.
Open Education level fullscreenThe SALSA map — who lands above expectations?
The SALSA model calculates an expected merit score from parents' education, the share of newly arrived pupils and the share of boys. The model contains no income variable, so the income differences in the first charts are not adjusted away here.
The deviation still tracks the SALSA data's own merit scores fairly closely (r ≈ 0.7), partly because it is calculated as actual minus expected value: municipalities with high merit scores usually land above zero, and the few with modest merit scores that land above expectations often rest on one or two schools. The municipal value is an unweighted mean of school deviations, and every municipality counts equally. Two in three municipalities fall below zero (median −5), mostly small ones with one or two schools, although the school mean is zero.
Green: results above expectations. Red: below expectations. Adjusted for the background factors included in the SALSA model.
Open SALSA map by municipality fullscreenThe data cannot show how much schools matter. The correlations apply to municipalities, not individual pupils, and they do not measure what schools do. The municipalities above expectations on the SALSA map are worth a closer look, but the map does not explain why they are there.
The pattern shows that school results co-vary with conditions outside school. That is important to keep in mind when schools and municipalities are compared with each other.
What the data does not show
- Municipal averages say nothing about individual pupils or schools.
- Income refers to 2024, education level to 2025, merit scores mainly to the 2025/26 school year and SALSA to 2024/25. The measures therefore refer to different years.
- The municipal values are unweighted means of schools, and many municipalities have only one to three schools with a merit score. Rank correlations are weaker than the stated r values, so municipalities with extreme values drive part of the correlations.
- SALSA adjusts only for parents' education, the share of newly arrived pupils and the share of boys.
- Correlations between municipalities at one point in time do not show cause and effect.