Crime Paywall free

Why Sweden doesn't keep stats on ethnicity and crime

Lee Roden
Lee Roden - [email protected]
Why Sweden doesn't keep stats on ethnicity and crime
File photo of a police cordon in Sweden. Photo: Johan Nilsson/TT

The subject of immigration is an increasingly charged one, yet much of the discussion is founded on guesswork. To the ire of some, and the agreement of others, Sweden does not collect statistical data on the ethnic background of criminals. The Local's Sweden in Focus series looks at why that is the case, and whether it's likely to change soon.


Contemporary debate about immigration in Sweden often centres on the unusually high influx of refugees in the last decadewhich is frequently claimed by the international far right to have sown chaos. Google's data shows that some of the searches connected to "Sweden immigration" that have increased most in the last five years are "Sweden crime rate" and "Sweden rape statistics". 

While the domestic debate is more nuanced, migration remains a hot topic in the build-up to the forthcoming autumn general election, and crime in relation to immigration is a particularly big talking point.

It doesn't take much effort to find articles claiming to have the 'truth' about the relationship between immigrants or refugees and Swedish crime. Yet the real truth is that there is no up-to-date public data on the ethnic background of criminals in the country, with existing figures more than a decade old.

There was once a time when Sweden regularly recorded such data on the national origin of those involved in crime, explains Stockholm University criminology professor emeritus Henrik Tham, who specializes in Swedish criminal policy and its history.

"In the 1980s in official statistics there were two tables on people who had been prosecuted, and one of them covered foreign citizens registered in Sweden under the category 'Living in Sweden'. There was a separate part about those not registered in Sweden: everything from students to Danes and Norwegians who came over and sometimes ended up involved in fights, but didn't have residence here," Tham tells The Local.

"In 1991 when the centre-right government decided it wanted to get rid of Statistics Sweden (SCB) – which was seen as an expression of social democracy – they moved the responsibility for statistics to different bodies, which meant crime statistics were moved to the National Council on Crime Prevention (Brå). At that point the stats we spoke about earlier disappeared."

SWEDEN IN FOCUS: An in-depth look at what makes this country tick

Stockholm University professor emeritus Henrik Tham. Photo: Stockholm University

Brå took a different approach to measuring statistics in the area, and instead of publishing regular figures on immigrants and crime opted for more sporadic but in-depth reports on the matter. The first of them, "Crime among immigrants and their children – a statistical analysis" (Invandrare och invandrares barns brottslighet – en statistisk analys) covered the period between 1985 and 1989, and was published in 1996. The report met with some controversy, as by the mid-90s attitudes in Sweden had changed.

"The interesting thing is that in several previous instances there were reports written which detailed the matter of crime by foreigners – I was involved in two of them – and at that time there was no opposition. But when the Brå report was published in the 1990s, people asked how long people were going to be treated as immigrants in connection with the stats – two, three, four generations? There was a lot of discussion about that," Stockholm University's Tham recalls.

Sensitivity about detailing ethnicity in relation to crime in Sweden only grew as the 90s progressed to the new millennium. Brå's last report on the matter, "Crime among persons born in Sweden and other countries" is now more than 10 years old. Stina Holmberg was one of the authors.

"The judgement we made is that looking at the four previous studies we did of roughly the same kind the results were very similar. Immigrants were overrepresented in crime, but the gap was reducing by the generation. Second generations are less over-represented. So we felt that there was no huge value in doing more because the results were pretty stable – what we knew was sufficient for our work," Holmberg explains to The Local.

Brå's first study analyzed data up to and including 1989, but significant changes had occurred by the time the 2005 report was compiled – including the arrival of almost 300,000 refugees in Sweden spearheaded by the fall-out from the Yugoslav wars. Immigration in general had also increased, while on top of that came the fallout from the recession in the country during the 1990s which increased segregation.

IN DETAIL: What lessons can Sweden learn from its Yugoslavian refugees?

Brå judged that the results of their new report were stable compared to past editions, despite those significant developments in scale of immigration and the economic situation. The vast majority of immigrants were not registered in connection with crime, but the risk continued to be higher than it was for those born within the country, and was also higher for those who have one or more foreign-born parent.

The 2005 report showed that those born abroad were 2.5 times more likely to be registered as a crime suspect than a Swedish born person with both parents born in the country (up from 2.1 in 1996). For those born in Sweden to two foreign-born parents that number decreased to two times as likely to be registered as a suspect, and for those with one Swedish-born parent it dropped to 1.4.

According to Brå in their report "if one standardizes in order to take into account differences in relation to age, gender, level of education and income, the relative risk among those born outside Sweden diminishes to 2.1". In other words, the same level as their first report.

"If the results are more or less the same, what do we do with them? People who are for immigration will say 'there's only slight change because of immigration'. Those who are against it will say 'look, I told you!'. These stats are always used as part of arguments on immigration rather than as a real method of measuring crime," Holmberg argues.

Brå researcher Stina Holmberg. Photo: The Local

Brå's position has been for some time that the 2005 report is still relevant, and that they have already effectively measured the crime-related consequences of periods of high immigration. 

Stockholm University's Tham fundamentally agrees that the results of research in the field have been consistent over the years:

"Several studies were done and the results were similar – immigrants were more likely to be involved in crime, but their children were less likely than they were."

He does however think that the older stats recorded in the era before Brå was tasked with detailing crime were still useful.

"It's a political question more than a research question, but if you look back 25 years, you could argue it's better to have kept the stats we had. I don't think that would have been enough to counter the statements coming from some groups today though, because it was a different discussion about migration then."

There are however some less talked about consequences of the reluctance to keep data on ethnic background in connection to crime. While much of the contemporary debate focuses on whether immigrants commit crimes or not, what about measuring how they are treated by law enforcement?

In the UK for example, official stats have helped to expose the disproportionate amount of ethnic minorities stopped and searched by police. In Sweden, the lack of official numbers means analysis in that area is a far more difficult task. Criminologist Leandro Schclarek Mulinari, who wrote a report on ethnic profiling by police in Sweden for NGO Civil Rights Defenders, says it's a frustrating situation.

"We lack a lot of stats. When you compare us to the studies in the USA or the UK for example, where there are stats involving ethnicity or race, you can see there clearly what the experience is for someone of a particular race or ethnicity. Those stats speak for themselves – that there's a disproportionate number of people of certain races stopped and searched by the police," he tells The Local.

"There are no stats similar to the stop and search stats according to race which exist in the UK. The small studies we have are very limited too. There are around 20 studies since the 1970s looking at whether immigrants are over-represented in crime stats, but when it comes to the other question, about whether they're discriminated against in law enforcement, there's only very small, limited studies. It's a problem that we don't have numbers."

The researcher thinks an explanation is that discussions on race and racism are not as advanced in Sweden as they are in countries like the UK or US, where mass immigration has been going on for much longer.

"That's to do with history of course – in the UK there has been migration from the Caribbean, for example, decades ago, and the discussion starts way back then. That also informs discussions about institutional racism. In the USA there has been a discussion for a long time too. In Sweden the first scientific study in this area happened in the mid 1990s. It's only 20 years we've been talking about this."

"There's also this idea that Sweden is an anti-racist country. This self-understanding that racism doesn't exist – that we're colour blind," he adds.

But could Sweden's approach to ethnicity and crime statistics be about to change? The current Social Democrat government has said further statistics in the area are not needed. The centre-right opposition Moderate party disagrees, saying studies should be commissioned detailing the ethnic background of perpetrators of sexual crime in particular, in response to an increase in reported rapes in the country as well as an increase in the proportion of people who were victims of sexual assault since 2015.

READ ALSO: Growing number of people in Sweden are victims of sexual crime

Tomas Tobé. Photo: Lars Schröder/TT

"The Moderates want to do as broad and comprehensive a study as possible on sexual crime, which has increased markedly in recent years," Moderate party secretary Tomas Tobé tells The Local via e-mail.

"That could for example be regarding the age of the perpetrator and their relationship with the victim. It could also be on whether the perpetrator has a foreign background. Such information could be useful for both changing legislation as well as shaping information efforts and other measures aimed at preventing crime. There could be grounds for also looking at other kinds of crime from that perspective." 

Asked about the argument made in some quarters that the stats could be manipulated for populist means, Tobé countered that knowledge is better than no knowledge when it comes to preventing myth-building.

"Not having sufficient knowledge when it comes to these questions is a problem, which risks contributing to false ideas and incorrect action. When it comes to sexual crime, we have called for as broad a mapping as possible in order to maximize knowledge. In addition, more knowledge can also help to prevent populist myth building from spreading – which Brå has previously noted," he replies, referring to Brå's justification for carrying out its 2005 study on crime among people born in Sweden and other countries.

It's little surprise that the likes of the anti-immigration Sweden Democrats back the Moderate calls for fresh stats on ethnicity and crime, but general opinion also seems to be shifting as it's not only the right who want to see more detailed figures. At their February party conference the Left Party backed the gathering of statistical data on ethnicity as part of anti-discrimination measures. Leading Swedish daily Dagens Nyheter meanwhile, which has a generally liberal stance, published an editorial last year backing the mapping of crime in relation to people with a foreign background.

"'We already know' is a weak argument," the editorial contested, adding that it is difficult to tackle problems and come up with adequate countermeasures without the availability of sufficient facts.


While we have your attention… a small request.

The Local has been giving you the story of Sweden for the past 14 years. But here's the thing: advertising revenues across the media industry – the revenues that keep us going – are more unreliable than ever.

That's why we're asking readers to become Members. For less than the price of a latte a month you can get access to all The Local's articles, including articles written just for Members. We'll also invite you to special events and give you exclusive offers. At the same time you'll ensure The Local stays thriving into the future.

Become a Member here. Or find out more here.


Calls for more thorough statistics were again heard after the Aftonbladet tabloid this week published a report on gang rape, finding that out of 112 people found guilty of gang rape in 2012-2017, 82 were born outside of Europe.

Brå's report from 2005 found that people suspected of rape or attempted rape were five times more likely to be foreign-born (however less than 0.3 percent of foreign-born people were suspected of such offences).

And on Tuesday Brå said it would look into whether or not it would re-examine the issue in a new report.

"It can now be established that the public demand for updated information on the relationship between crime and descent and migration is so great that the authority needs to consider the possibilities again," it said in a statement to the TT news agency just days after the interviews for The Local's article were conducted.

Brå's reports are planned far in advance and it said it would not make a decision until early next year.

Before Tuesday's statement, Brå's Holmberg told The Local she understood the argument that new numbers are better than no numbers if speculation is to be lessened, but did not necessarily agree with it. 

"If the government tasked us with doing that, I'd understand the argument. As a knowledge-based agency, we don't feel like there's a lack of knowledge that's so big as to require us to do it. There are enough valuable studies."

"I felt that when we worked on it the last time, it didn't improve the quality of discussion in the area compared to the edition before – it was still people in favour of immigration showing it proved their point, or people against saying it proved theirs," she adds.

Criminologist Schclarek Mulinari points out that the discussion on future steps should ultimately not only be about producing more stats, but also about how they are framed and acted upon.

"More often than not the questions are framed in a way that's related to being repressive towards immigrants. It's worth remembering that numbers have to be interpreted, and it's also about what kind of political interest there are in using the stats, how those stats are read. Stats are in themselves neutral, they have to be interpreted," he emphasizes.

"If you really want to deepen understanding in this area you have to consider what the stats are and also what they're going to be used for. Are they going to be used to better locate people without residence papers, for example? To track down migrants and their kids? Or are they going to be used to improve how the police act, and to create a less discriminatory police? It's very complicated, and more often than not the discussion in the area in Sweden is too simplistic," he concludes.

This feature is part of The Local's Sweden in Focus series, an in-depth look at what makes this country tick. If you're a Member of The Local, why not share your thoughts in the comments below. If you're not yet a Member, click here to join.


Join the conversation in our comments section below. Share your own views and experience and if you have a question or suggestion for our journalists then email us at [email protected].
Please keep comments civil, constructive and on topic – and make sure to read our terms of use before getting involved.

Please log in to leave a comment.

See Also