What percentage of immigrants are supposed to be committing rapes in Sweden
Executive summary
Recent Swedish research and investigative reporting show that a clear plurality — and in some studies a majority — of individuals convicted of rape have an immigrant background, but the exact percentage varies by dataset and definition: major analyses report figures around 58–63% of convicted offenders having immigrant or foreign-born backgrounds [1] [2] [3], and rigorous case–control work confirms an overrepresentation of people with an immigrant background even after statistical adjustment for socioeconomic and clinical confounders [4] [5].
1. What the headline numbers say: ~58–63% in prominent counts
National and academic tallies commonly cited in Swedish debate find that between about 58% and 63% of those convicted of rape or “rape+” (rape, aggravated rape, attempted rape) in particular samples have an immigrant or foreign-born background: Swedish Television and BBC reporting relayed that 58% of men convicted in a recent five‑year window were born abroad [1], a latent‑class offender study found 59.2% of 3,039 convicted offenders were immigrants (47.7% born outside Sweden) [2], and coverage of the Lund University case–control study has been summarized as roughly two‑thirds (about 63%) of convicted rapists having an immigrant background [3] [6].
2. The largest recent academic study: design and headline finding
A 21‑year Swedish case–control follow‑up analyzed 4,032 individuals convicted of rape+ and 20,160 matched controls and reported a strong association between immigrant background and rape convictions that persisted after adjusting for socioeconomic status, substance use, psychiatric disorders and prior criminality, with the effect especially marked for those born abroad who arrived as teenagers or older; the study’s sample composition and matched design are described in the published article and institutional summaries [4] [5] [7].
3. Important definitional and methodological limits that change interpretation
These percentages describe convicted offenders in specific samples and time windows, not the share of all immigrants who commit rape; convictions reflect reporting rates, policing, prosecutorial decisions and legal definitions, and Sweden has changed rape legislation and recording practices that affect counts over time [8] [4]. Academic authors and commentators explicitly call for caution: overrepresentation in conviction data does not equal causation and the mechanisms behind the association (integration, acculturation, socioeconomic factors, age at arrival) remain under investigation [6] [5].
4. Confounding, nuance and alternative measurements
Researchers reduced, but did not eliminate, the association by adjusting for confounders, meaning socioeconomic disadvantage, mental‑health and substance problems, and criminal history explain part of the overrepresentation but not all of it according to the Lund University team [4] [5]. Other work highlights how juvenile crime trends and changing demographics complicate simple comparisons, and some authors note that proportions can shift depending on whether second‑generation immigrants are counted as having an “immigrant background” or not [9] [2].
5. What this does and does not prove — and what reporting agendas matter
The empirical record summarized here proves that convicted rape offenders in several Swedish datasets are disproportionately likely to have immigrant backgrounds by the percentages cited (around 58–63% in high‑profile studies and reports) [1] [2] [3], but it does not prove broad cultural determinism or that most immigrants commit sexual violence: the data are convictions, not population prevalence of offending, and researchers explicitly call for further work on mechanisms and policy‑relevant causes [6] [5]. Media and political actors have used these figures for differing agendas — from calls for deportation to demands for improved integration and social services — so the numbers must be read alongside methodological caveats already acknowledged by the researchers and reporting outlets [1] [6].