Last updated: 2026-10-19

Immigration and crime: what the data show in the US, Nordic countries, France and the UK

Official statistics count different things in each country: registered suspects (Sweden, Finland, France), charged persons (Norway) or persons found guilty (Denmark), grouped by birth country, parents' birth country or nationality. Where adjusted figures exist, as in Sweden and Norway, the gap between groups narrows after allowing for age, sex, income or employment but does not close. France records nationality but not ethnicity, the UK and US sources contain no comparable origin series, and no figure here measures how many people in a group commit crime or why.

This page sets out what each country records, shows unadjusted and adjusted figures side by side, and lists what does not exist. The detailed Nordic tables in the sources cover one origin group, people born in or of Somali origin, because that is where the agencies' country-level tables were retrieved. They are not a verdict on all immigrants. Registered suspects, charges and convictions are not the same thing as offences committed. Every figure links to its source.

What does the data show about immigration and crime?

The data show that results depend on the measure, the country and the origin group. The table summarises what the sources behind this site cover.

The three stages of the criminal process differ. Sweden's Brå and Statistics Finland count suspects, Statistics Norway counts charged persons, and Statistics Denmark counts persons found guilty. A figure from one stage cannot be compared with a figure from another.

What does Sweden record about suspects by origin?

Brå's 2021 report counts persons aged 15 and over living in Sweden on 31 December 2014 who were registered as suspected of at least one offence between 2015 and 2018. Of 45,218 Somalia-born people, 6,440 were registered as suspects. The reference group is people born in Sweden with two Swedish-born parents.

Three cautions apply. Registration as a suspect is not a conviction. The share is cumulative over four years. The authors describe the study as mainly descriptive, not causal. Because 14.24% were registered, about 86% of Somalia-born adults were not registered as suspects at all (site calculation). The annex covers suspects from 2015 to 2018 only.

What does Norway record about charges?

Statistics Norway's 2017 report counts persons charged in 2010 to 2013 per 1,000 residents aged 15 and over. For Somalia-born residents, the unadjusted rate was 123.8 per 1,000, against 44.9 for the rest of the population, based on 1,800 persons.

The report notes that over-representation had fallen since an earlier SSB study, which found 227 per 1,000 for Somalia-born residents, although the two studies are not strictly comparable. In this group, adding employment reduced the gap more than adding age and sex alone.

A newer SSB article on charges in 2020 to 2023 reports that men aged 15 to 24 from Iraq and Somalia had charge rates close to 1,200 per 1,000 over four years, and Norwegian-born men with Moroccan or Somali parents close to 1,300, against just over 280 for the rest of the population. SSB cautions that these are charges, not persons, that a few heavily charged individuals can move a group's rate, and that only the rates for men aged 15 to 24 are comparable by age and sex.

What does Denmark record about convictions?

Statistics Denmark publishes persons found guilty by country of origin, including minor and traffic offences. The rates below are this site's calculations from STRAFNA3 counts and the FOLK1C population.

Of the 1,808 Somali-origin persons found guilty, 1,560 were men and 248 were women. The age-specific rows are the nearest published substitute for age standardisation, and they do not adjust for income, education or employment. For sexual offences the 2025 counts were 9 Somali-origin persons and 839 Danish-origin persons, too few on the Somali side to turn into a rate. Statistics Denmark's age-standardised index was not retrieved.

What does Finland record about suspects?

Statistics Finland publishes suspects of solved Criminal Code offences by citizenship, so naturalised Finns of Somali background are not in the Somali row. In 2025 the rate for Somali citizens was 2,471.9 per 10,000 (1,714 persons), against 602.8 for Finnish citizens.

All foreign citizens combined had 812.2 suspects per 10,000. The Somali citizen rate was 1,988.6 in 2023, 2,357.7 in 2024 and 2,471.9 in 2025, and the small base means large yearly swings. Because the series is by citizenship, it is not comparable with the background-based figures for Sweden, Norway and Denmark.

How do the Nordic figures compare side by side?

The units differ, so compare rows within a country, not across countries.

In Sweden and Norway the ratio falls once controls are added. In Denmark and Finland the age-specific rows show gaps within the same age and sex band, but those sources do not adjust for income or employment. The Netherlands is the other country checked: CBS publishes suspect rates for only a few named origin groups and none for Somalia.

What does France record, and what does it not?

France does not record ethnicity or religion in official files. Administrative data cover sex, age, nationality, place of birth and criminal record. The figures below therefore describe nationality only.

A parliamentary question citing 2024 data, which this site has not checked against the original, reports that foreign nationals are about 8% of the population but 18% of people suspected of homicide, 14% for rape and attempted rape, 37 to 38% for burglary and 22% for drug trafficking. In transport settings the foreign share of suspects is far higher: 80% of non-violent thefts against 26% elsewhere. The same source says department- and commune-level shares are unreliable.

For the 2023 riots, the justice and interior inspectorates found that 79% of suspects and 78% of 395 convicted adults held French nationality, and 75% of suspects were born in France. The Interior Minister gave 90% French among those arrested in July 2023, so sources differ. Descent is not recorded; one Paris police note described most arrested rioters as young French nationals from immigrant families, but the method is unpublished, so it is an assessment, not a statistic. See the 2023 riots page.

For the 17 and 18 October 2026 riots, no nationality breakdown of the 2,397 arrests had been published by 19 October, so any claim about the share of foreigners is unsupported for now. Recorded crime overall counts reports and police activity, not actual crime; the Interior Ministry's 2025 first look found property offences mostly falling and offences against persons still rising, more slowly than the ten-year average.

What do UK sources show?

The sources reviewed found no Home Office or Ministry of Justice publication of crime by Somali origin, and no official origin series was retrieved for the UK. The Office for National Statistics does publish Census 2021 data on 108,921 people in England and Wales born in Somalia and 176,645 who identify as Somali, but those tables cover population, work and housing, not offending. A freedom-of-information analysis by an advocacy group was cited in earlier research but not verified, so it is not used.

UK disorder counts exist but do not record origin. After the Southport stabbings, 1,840 arrests and 1,103 charges had been recorded by July 2025, according to a secondary summary. In the summer 2025 Epping hotel protests, 28 people were arrested and 16 charged in one police tally. See the United Kingdom page.

What do US sources show?

The US sources behind this site contain no national study comparing offending by nativity. They contain enforcement records and case counts, each with its own definition.

Detainees with criminal records. ICE says about 60% of 65,765 detainees had a criminal record in July 2026. TRAC reported that 70.8% of the 60,311 people detained on 4 April 2026 had no criminal conviction. The measures differ: one counts records including pending charges, the other counts convictions. Analyst Austin Kocher reported that 92% of fiscal 2026 detention growth came from people with no criminal convictions.

Minneapolis-St. Paul arrests. Freedom of Information data analysed by the Deportation Data Project show at least 3,789 arrests. Of the arrests between 1 December 2025 and 10 March 2026, Somalia-born people accounted for 106, under 3%. Fewer than a quarter of those arrested had criminal records, and DHS disputes the dataset.

Minnesota fraud cases. The Justice Department says Feeding Our Future fraudulently obtained more than $240 million, and 78 defendants were charged, 66 pleaded guilty or were convicted as of May 2026. Across all Minnesota program-fraud cases, Attorney General Pam Bondi said 98 individuals were charged, 85 of Somali descent. Eighty-five people is about 0.08% of the roughly 107,000 Minnesotans of Somali descent cited in press accounts (site calculation). Fraud requires access to a program, so charged persons are not a random sample, and the percentage estimates nothing about a community. Descent is not immigration status: most of those charged are US citizens. See Minnesota fraud cases.

Terrorism cases. The cases are individual charges: about 20 men in the 2007 to 2009 recruitment wave, nine in the 2015 to 2016 Islamic State case, one attacker in 2016, and one 18-year-old arrested in October 2026, against a Minnesota community of 76,320 by ancestry in the 2024 survey and about 107,000 to 108,000 on the wider press-quoted count. None is a charge against a community.

Gang claims. After a shooting on 2 October 2026, a law-enforcement statement about regional gang violence was reported by MPR, which found no published data behind it. Another MPR article noted it offered no police data on gang membership for campaign-ad claims.

Hate-crime data. Minnesota's Bureau of Criminal Apprehension recorded 224 bias-motivated incidents in 2025 and 228 in 2024, with no published breakdown for Somali victims.

What crime data does not exist?

Several gaps limit what can be said.

• France: ethnicity and religion are not recorded, so no claim by descent can be measured.
• United Kingdom: no official publication by Somali origin was found.
• Netherlands: no Somalia-specific series.
• Sweden after 2018: the Brå annex stops at 2018.
• Norway 2015 to 2018: SSB's charge tables were commissioned by a parliamentary party, are not in StatBank, and have no Somalia row in the retrievable text.
• Germany and Italy: no crime series by Somali origin was retrieved.
• Jihadist offenders: no European official statistic counts them by country of birth, so only individual court cases can be cited.
• United States: no national comparison by nativity appears in the sources, and Census data do not record whether a person is authorized to be in the country.

"Not found" means not located in this research, not that no data are held.

How should immigration and crime figures be read?

Six features of the data change what a figure means.

1. Age and sex structure. The Brå figure falls from 14.24% to 9.03% once age and sex are applied, and in Denmark the Somali-origin rate is 20.8% for men aged 15 to 29 and 3.1% for women of the same age. France's inspectorates found that 18-to-24-year-olds were 8.28% of the population but 73% of the riot sample.
2. Income, education and employment. Controls for income and employment lowered the Swedish and Norwegian figures, as the tables above show. Somali-origin employment rates in the sources were 51.1% in Denmark and 48.8% in Norway, against 79.7% and 79.5% for the majority groups.
3. Policing intensity. The National Assembly source lists police targeting among the factors that affect the foreign share of suspects.
4. Reporting. Recorded crime counts reports and police activity, not actual crime, and victim surveys differ.
5. Who is in the denominator. Nationality series omit naturalised people, and the French source notes that irregular migrants and transit populations such as tourists can raise foreign shares. The French source also notes offences that only foreigners can commit.
6. Small numbers and heavy offenders. SSB warns that a few heavily charged individuals can move a group's rate. Finland's Somali citizen rate moved from 1,988.6 to 2,471.9 in two years.

What are the competing interpretations?

The sources attribute different readings to named bodies, and this page does not choose between them.

• Descriptive, not causal. Brå's authors describe their study as mainly descriptive, and SSB shows employment reduces the gap more than age and sex alone for Somalia-born residents. These findings show that the measured gap changes with the controls used, though neither source assigns a cause.
• Persisting gaps. The Danish and Finnish age-specific rows show differences remain within the same age and sex bands. Those sources do not test explanations.
• Enforcement framing. The administration describes its US operations as targeting "criminal illegal aliens". TRAC and Kocher, using conviction records, report that most detainees have no conviction. ICE counts records including pending charges. The definitions differ, so the sources conflict without either being shown wrong.
• Fraud cases. Critics, including Republican legislators and the Center for Immigration Studies, argue that community networks and leaders did not do enough to prevent or expose fraud. Governor Walz and Somali leaders reply that most Somali Minnesotans have no involvement, that fraud is a crime wherever it originates, and that the ringleader of the largest case is white and American-born according to press descriptions. PolitiFact rated Bondi's claim "Half True" because it implied the effort was recent.
• French suspects. The parliamentary question reports the shares above, and its own caveats say suspects are not convictions and the figures describe nationality, not any ethnic or religious group, and do not show cause.

Does tighter immigration policy change crime?

The sources do not establish a causal link either way. Sweden's government came to power in 2022 pledging to cut gang crime and immigration, and Brå recorded 84 gun-violence deaths in 2025, down from 92 in 2024 and 124 at the 2020 peak, with one February 2025 mass shooting accounting for 10 deaths. Officials credit surveillance powers, tougher sentences and more resources. Sweden's foreign-born share is 20.8%, against 14% in France. Countries that tightened policy, including Denmark, Sweden, Germany, Italy and the Netherlands, show different outcomes on their own measures. See the Europe page.

Related pages

• World migration overview
• Immigration by country: 15-country comparison
• Somalis in Europe
• United States immigration
• France immigration
• Frequently asked questions
• Sources and method