RUTHFULLY YOURS

Last Saturday, another controversy erupted involving the now familiar mix of President Trump, the media, and immigration.

In a speech, Trump riffed on a Fox News segment he’d seen on immigration and crime in Sweden, causing much confusion. The situation hardly improved as journalists and pundits mostly unfamiliar with the topic rushed to explain the finer points of Swedish crime statistics.

So, what is the situation actually like? Nightmarish rape capital of Europe or a safe welfare state unfairly maligned by far-right agitators?

Sweden has a growing problem with crime that is linked to immigration, but the Fox News segment was sensationalistic. As with many exaggerated reports from Sweden in foreign right-wing outlets, the tone of the reporting implies there has been a large crime wave brought about by the recent migration crisis. This is misleading.

Refugees who arrived during the migration crisis are too few in number to explain much of Swedish crime trends. Sweden’s crime-heavy immigrant neighborhoods emerged gradually through the accumulated effects of many decades of immigration.

Several types of crimes such as gang shooting, arson, and sexual assault have increased in Sweden, but other categories such as assault, car thefts, and property crimes have decreased. The increase in sexual assault and violent crime is not as spectacular a development as the Fox News segment made it out to be. Even in Swedish immigrant enclaves, criminality is still fairly mild compared with U.S. crime hubs. Last year the famously multicultural Swedish city of Malmö had a homicide rate of 3 per 100,000, far lower than the 28 per 100,000 rate in Chicago.

In their response to Donald Trump, the Swedish government has pointed out that the homicide rate in Sweden is lower now than in 1990. We should nevertheless note that the homicide rate has decreased in almost every Western country since 1990, owing to social reasons, changes in attitudes, and, in part, medical advances that save the lives of more crime victims. The homicide rate in Sweden has declined less than in the United States, Western Europe, and other Nordic countries, and has increased again the last few years.

Between 1990 and 2015, the homicide rate in Sweden declined from 1.3 to 1.1 per 100,000. This drop is less than that in Western Europe as a whole, where the homicide rate declined from 1.3 to 0.6 in 2013, the latest year reported by the World Health Organization.

In Finland, the homicide rate declined from 3.2 to 1.3 during the same period, and in Norway from 1.1 to 0.4. The rate was stable at 0.8 in Denmark.

While the homicide rate inevitably varies in a small country year by year, Sweden appears to have transformed from one of the lower-crime countries in Western Europe to above average. We cannot say for certain how much immigration contributes to violent crime in Sweden. The numbers are collected by statistical agencies in Sweden, but they have not been reported since 2005 because of the informal taboo on linking immigration to crime.

While Fox paints too dark a picture of Sweden, other outlets are going to the other extreme in their eagerness to refute Donald Trump. The Washington Post credulously reports some highly questionable claims made in the Swedish newspaper DN.

The Washington Post cites criminologist Felipe Estrada Dörner: “Overall, Sweden’s average crime rate has fallen in recent years, Dörner said. That drop has been observed for cases of lethal violence and for sexual assaults.”

The associated graph shown by the Washington Post appears to confirm this picture for recent years, with data based on the self-reported NTU survey of crime victimization (the acronym comes from the Swedish name of the national survey).

This is either incompetence or manipulation. The text in the Washington Post discusses “sexual assault,” which spiked recently in Sweden in the NTU survey. The graph instead displays instances of “assault,” a non-sexual crime.

In quantitative terms, the NTU data on sexual assault show a rising victimization level from the first survey in 2005 to the last reported year, 2015. The legal definition of rape in Sweden, suddenly a hot topic in U.S. political discourse, is irrelevant here. These numbers are from a self-reported survey of sexual-crime victimization with the same definition over time. The number of sexual crimes reported to the police has also increased in recent decades, although here crime definitions and reporting rates become factors and make comparisons over time and with other countries difficult. There is no source pointing to a decrease in sexual crime over the long term in Sweden, and the increase in sexual assaults in official sources was grudgingly acknowledged by the Swedish media prior to Donald Trump’s comments.

Finally, it is worth emphasizing that the focus on overall crime statistics avoids the actual question being discussed: the situation in immigrant-dominated areas with low socioeconomic status. Around 95 percent of the Swedish population lives outside these areas. Hence, it takes a lot for crime trends in the “ghettoes” of Sweden to dominate overall trends. Four fifths of the population in Sweden are not immigrants. The overall crime trend and the effect of immigration on crime are therefore two different topics.

To isolate the effect of immigration on crime, we need data on crimes committed by immigrants. Obtaining this type of data is easy in the United States or Denmark, but not in Sweden. The last time there was an official report breaking down crime statistics by immigrant status and origin was in 2005, for the years 1997 to 2001. These statistics confirmed that immigrants were significantly overrepresented amongst offenders, in particular in committing violent crimes. The foreign born were four times more likely to be suspects in homicide cases than those with Swedish origin, and 4.5 times more likely to be suspects in rape cases.

Since then, Swedish criminologists and politicians have made sure that no new statistics have been released. Not a single recent research study in Sweden has attempted to estimate the causal effect of immigration on sexual assault or homicide rates. Parliament recently defeated a motion to produce up-to-date crime statistics based on national origin. We simply do not know what percentage of sexual assaults or homicides were committed by immigrants last year in Sweden. The Swedish criminologists and government officials who adamantly deny the effect of immigration on crime don’t know these figures, and strikingly don’t want to know. Americans who are interested in this topic should focus on this surreal taboo against statistics, not cartoonish exaggerations that falsely portray Sweden as a war zone.

The commitment to secrecy is a perfect example of how Western governments are fueling populism and distrust in established politicians. In order to get a less Trump-based and more fact-based debate, reasoned discussion based on current statistics would be a good start.

— Tino Sanandaji has a Ph.D. in Public Policy from the University of Chicago, and is currently a researcher at the Institute for Economic and Business History Research in Stockholm.