Utländsk överrepresentation i grov brottslighet

What does the data show?

Objective observations — not interpretations

• Brå 2021:9 is the most recent comprehensive register study of crime suspects by background. It draws on the whole population 2007–2018 (not a sample) and standardises for age and sex — two factors that otherwise explain much of crime patterns in all countries.
• After standardisation, foreign-born persons are 2.5 times more likely than natives with Swedish background to be suspected of serious violence. The second generation (native-born with two foreign-born parents) has around 1.8× overrisk — higher than the first generation from comparable regions in some cases.
• Overrisk varies sharply by region of origin. From the Middle East and North Africa the overrisk is 2.5–3× for murder/manslaughter; from Eastern European EU countries and Southeast Asia it is close to 1.0×. This is documented in Brå 2021:9 table 5.4.
• Absolute share and relative risk are different things. About 53 % of all suspects for serious violence have a foreign background (Brå 2021:9) — but since the group is 38 % of the population, the relative risk is ×1.4 without standardisation, ×2.5 after standardising for age and sex.
• For fatal shootings the overrisk is higher than for any other crime (×3.5–5 per Sturup et al. 2019 and Swedish Police NOA 2024). Brå has not published a separate overrisk figure for fatal shootings — the research-based estimate is shown here.
• No new Brå study after 2018 has been published despite a government commission delivered April 2024. The most up-to-date national figure is thus seven years old — a central methodological limitation for all current debates.

Definitions

How the numbers are counted — and what they do not cover

Overrisk (relative risk)

The ratio between the probability that a person in group A is suspected of a crime and the same probability in group B. Expressed as ×N, where ×1.0 = no difference. Overrisk does not say how many crimes a group commits in total — only how much more common it is per person in that group.Normally standardised for age and sex, since these are strong background variables for crime in all countries.

Foreign background (Swedish definition, SCB)

Person who is foreign-born OR native-born with two foreign-born parents. Narrower than several EU countries' definitions (which often count one parent). As of 2024, the group is about 38 % of Sweden's population.

Foreign background (Danish definition)

Person who is »invandrer« (foreign-born) OR »efterkommer« (native-born with at least one foreign-born parent). Under the Danish definition applied to Swedish data, the overrepresented group would be larger since more people are counted.

Standardisation (age and sex)

Statistical adjustment that removes the effect of groups having different age structure and sex distribution. Without standardisation, a younger group would automatically show higher overrisk for crime since crime peaks at 15–29.

Absolute share of suspects

The share of all suspects belonging to a given group — e.g. »53 % of serious violence suspects have a foreign background«. This is not the same as overrisk and can be misleading without comparing to the group's population share.

Second generation

Native-born person with two foreign-born parents. Brå 2021:9 shows that the second generation often has HIGHER overrisk than the first generation from the same region — a pattern seen in many other immigration countries.

Register data

Data based on public authorities' full registers (population records, LISA database, suspect register). Advantage: covers the whole population, no missing-data issues. Disadvantage: measures only what is recorded — dark figures and selection effects in policing affect the results.

Brå 2021:9

»Suspects for crime among persons with domestic and foreign background« — Swedish Crime Prevention Council report published 2021. Draws on the suspect register 2007–2018 linked to LISA. Most recent comprehensive register study of crime suspicions by background.

Brå 2005:17

Predecessor to 2021:9 — covers 1997–2001. Overrisk was then somewhat lower than in 2021:9 for most crime types. Provides a reference point for assessing change over time.

Sturup et al. 2019

Research study of fatal shootings in Sweden, published in the European Journal of Criminology. Provides the most cited estimate of overrisk for fatal shootings (×3.5–5). Based on Swedish Police data 2000–2017.

Frequently asked questions

Short answers to what is most often discussed