Analysis · 2 min read
Overrepresentation in criminal suspicion — what Brå2021:9 shows
Brå2021:9 measures suspicion of at least one offence in2015–2018 among people aged15+ registered on31 December2014. Ratios vary between groups, offence types and adjustment levels. Suspicion is not conviction and results do not establish causes.
Guiding question:What does Brå2021:9 show about criminal suspicion2015–2018 for different background groups and adjustment levels?
Short version in the chapter: Foreign overrepresentation — chapter
Published 2026-07-16
Quick stats
- Risk ratio all offences, foreign-born
- 2.25×
- Risk ratio robbery, second generation
- 4.7×
Brå 2021:9 fig. 7, age and sex, suspicion 2015–2018
Brå 2021:9 fig. 13b, age and sex
What is measured?
The outcome is a person being registered as suspected of at least one offence in2015–2018. The population is people aged15+ registered on31 December2014. The reference group is Swedish-born people with two Swedish-born parents, with ratio1.00.
Earlier Brå reports are historical background, not a continuous25-year series on this page. Unbound historical comparisons and prison shares are not used to explain this study’s results.
All offences: three adjustment levels
Figure7: foreign-born2.51 unstandardised,2.25 adjusted for age and sex,1.76 also adjusting for disposable income, own education and municipality type. Swedish-born with two foreign-born parents:3.21,1.96 and1.68. Swedish-born with one foreign-born and one Swedish-born parent:1.86,1.46 and1.34.
Second generation here means Swedish-born with two foreign-born parents. This group’s higher ratio than the foreign-born applies unadjusted; after age and sex it is lower. The chart shows only age-and-sex adjustment.
Exact chart observations
People aged 15+ registered on 31 December 2014. Reference: Swedish-born people with two Swedish-born parents. These ratios cover all offences, not a particular offence type or origin region.
Foreign-born
- Risk ratio, age and sex
- 2.25
Swedish-born with two foreign-born parents
- Risk ratio, age and sex
- 1.96
| Population | Risk ratio, age and sex |
|---|---|
| Foreign-born | 2.25 |
| Swedish-born with two foreign-born parents | 1.96 |
Figure7: the same outcome at three levels
People aged 15+ registered on 31 December 2014. Reference: Swedish-born people with two Swedish-born parents. These ratios cover all offences, not a particular offence type or origin region.
Foreign-born
- Unadjusted
- 2.51
- Age/sex
- 2.25
- Five factors
- 1.76
Swedish-born, two foreign-born parents
- Unadjusted
- 3.21
- Age/sex
- 1.96
- Five factors
- 1.68
Swedish-born, one foreign-born and one Swedish-born parent
- Unadjusted
- 1.86
- Age/sex
- 1.46
- Five factors
- 1.34
| Population | Unadjusted | Age/sex | Five factors |
|---|---|---|---|
| Foreign-born | 2.51 | 2.25 | 1.76 |
| Swedish-born, two foreign-born parents | 3.21 | 1.96 | 1.68 |
| Swedish-born, one foreign-born and one Swedish-born parent | 1.86 | 1.46 | 1.34 |
Offence types must remain separate
Figures13a–c distinguish all offences from specific offence types. For robbery, foreign-born ratios are3.5 unadjusted,3.5 after age/sex and2.0 after five factors; Swedish-born with two foreign-born parents:11.5,4.7 and3.1.
Lethal violence in these figures includes attempts. After five factors, foreign-born ratios are2.3 for lethal violence including attempts,2.2 for rape and0.9 for residential burglary. For Swedish-born with two foreign-born parents the corresponding ratios are3.4,1.4 and1.8. There is not one common ratio for all crime.
Method and limitations
The study measures registered suspicion, not convictions or all actual offending. It counts persons with at least one registered suspicion, not offences or a clearance rate.
The five factors are age, sex, disposable income, the individual’s own education and municipality type. Municipality type is not neighbourhood and income here is not parental income.
Standardised ratios describe associations. Remaining differences do not identify an ethnic cause and statistical adjustment is not causal analysis. Group averages establish no individual’s guilt, behaviour or ability.
How should results be used?
Compare only the same outcome, population, period and adjustment level. A group’s share among suspects and its population share is a different comparison from these risk ratios.
The material does not determine why a particular local safety trend occurs or which policy should be chosen. Earlier unsupported statements attributed to named people and supposed direct quotations are not used as research findings.
What this shows
- Figure7 compares suspicion of at least one offence2015–2018, not convictions. Age-and-sex ratios are2.25 for the foreign-born and1.96 for Swedish-born with two foreign-born parents.
- Differences after statistical adjustment show associations, not causes or individual guilt.
Definitions
- Risk ratio
- The group’s proportion with the registered outcome divided by the reference group’s proportion;1.00 means equal proportions.
- Second generation on this page
- Swedish-born with two foreign-born parents. The group with one foreign-born parent is shown separately.
- Five factors
- Age, sex, disposable income, own education and municipality type.
- Suspicion
- Registered criminal suspicion, not conviction.
Method and limitations
- The study measures registered suspicion, not convictions or all actual offending. It counts persons with at least one registered suspicion, not offences or a clearance rate.
- The five factors are age, sex, disposable income, the individual’s own education and municipality type. Municipality type is not neighbourhood and income here is not parental income.
- Standardised ratios describe associations. Remaining differences do not identify an ethnic cause and statistical adjustment is not causal analysis. Group averages establish no individual’s guilt, behaviour or ability.
Primary sources
FAQ
- Is suspicion the same as conviction?
- No. The outcome is registered suspicion of at least one offence2015–2018 in the study population, not convictions or offence counts.
- What happens after five factors?
- Figure7: foreign-born1.76 and Swedish-born with two foreign-born parents1.68 after age, sex, disposable income, own education and municipality type.
- Is the second-generation ratio higher?
- For all offences it is higher unadjusted:3.21 versus2.51. After age/sex it is lower:1.96 versus2.25. For robbery after age/sex it is higher:4.7 versus3.5. Offence type and adjustment level must be specified.
- Does this show an ethnic cause?
- No. Residual associations after statistical adjustment do not identify causes or establish individual behaviour or guilt.