We’ve all been reading quite a lot the last few days about “what is going on in Sweden” with conflicting reports indicating either that Sweden is a hotbed of crime and particularly rape, versus rebuttals claiming nothing much at all is going on in Sweden, move on please.
A couple of days ago, what started out with me becoming rather exercised over a post on Twitter, turned into a very useful exchange and eventually led me to do some research of my own.
The particular post that got my blood boiling (and the snarky tweet I sent in response) are shown below.
My response was directed at Chivers’ comment that “police-recorded rapes” ≠ “actual number of rapes”, in combination with a section from the underlying article that he had written, in which it was stated that while Sweden has a very high incidence of reported rapes, the number of actual rapes that take place is lower, one of the reasons being that the definition of “rape” is much broader in Sweden than in other countries, the other reason being that while every report is recorded as a crime, it may later transpire that there is insufficient evidence to prosecute, but that the incident will remain on the books as a crime.
Now, I accept that what may be classed as “rape” in Sweden may fall under some other category of sexual assault in other countries, but having had a quick browse through the sexual offences statistics on Sweden’s crime website, it was clear to me that Sweden certainly does not categorise all sex offences as “rape”, in fact in 2015 those offences which were classed as rape made up approximately 33% of sexual offences recorded by the police. So whatever their categorisation, I would argue that if it has been recorded as a rape then the underlying allegation is probably still fairly serious.
Whether or not there is sufficient evidence to prosecute (it’s well known that sexual assault cases are notoriously hard to prosecute due to difficulties in obtaining evidence), I would expect that Swedish women are no more likely to make false reports than those in any other western country. Hence my outrage at what I saw as the suggestion that the lower incidence of “actual rape” than “recorded rape” implied the statistics were not a cause for concern.
To his credit, Chivers immediately responded to my tweet to clarify what I was getting my knickers in a twist about (my words, not his) and after a long back-and-forth as to the relative definitions of sexual offence vs sexual assault vs rape, eventually I thanked him for his time and patience but was left with a lingering feeling that I still was no closer to understanding whether or not Sweden truly does have a problem with growing numbers of sexual assault.
A short while later, Chivers retweeted a truly excellent thread by Ciaran Jenkins, in which Jenkins very politely but very firmly took Nigel Farage to task over the same tweet that Chivers had done, specifically the references to the “UN graph”.
The key points made by Jenkins are reproduced below:
These tweets, while fascinating, only served to muddy the waters further in my mind – bearing in mind all these factors, how is somebody such as myself, who likes to deal in absolutes, supposed to make any kind of reasonable comparison?
Jenkins’ analysis, however, gave me the idea of conducting a comparison of the levels of sexual violence reported by the UNODC for just three countries – Sweden, Germany and the United Kingdom. All three countries have sexual violence laws in all three categories concerned, and all three countries score 3/3 on enforcement. Crucially, in the metadata that accompanies the statistics, all three countries state that the figures they have provided meet the UNODC’s definition of “sexual violence” which includes rape and sexual assault, as well as sexual offences against children
Source: unodc.org (link to full metadata here)
It struck me as a bit strange that Sweden states their “total sexual violence” data do not include attempts whereas data provided for rape and sexual offences against children do. I can only conclude that this means they record attempted rape and attempted child sexual offences but possibly not attempts within other categories of sexual offences (though I would not like to hypothesise as to what such a category would be).
The comparative figures for the three countries are shown below – figures are number of police-recorded incidents of sexual violence, per 100,000 population. It’s clear from the figures, as well as from the percentage increases that I have calculated in the far right-hand columns, that Sweden not only has had by far the highest absolute number of recorded incidents in recent years, but also the highest percentage increase from 2003 to 2014. On the other hand, comparing just the most recent period from 2010 to 2014, the highest percentage increase is to be seen in England and Wales, while the percentage increase in Sweden is relatively low. Germany has seen a decrease of 13% between 2003 to 2014, and almost no change between 2010 and 2014.
Source: unodc.org (link to download full dataset here)
The reason for the large percentage change in Sweden’s reporting figures over the 10 year period can be explained by a quote from Tim Newburn, professor of criminology at LSE, contained within Chivers’ original BuzzFeed article:
Given the relatively small increase in police-recorded incidents between 2010 and 2014, does this mean increases in sexual assault in Sweden are actually starting to flatten off in recent years? Sadly the UNODC does not yet provide data for 2015 or 2016 – for these years I went back to the Swedish website I had consulted earlier in the day. This site provides two separate sets of data – statistics regarding crimes reported to the police (which are the figures that are subsequently provided to the UNODC), and the annual Swedish Crime Survey.
The Swedish Crime Survey is described as “an annual survey of the attitudes and experiences of the general population of Sweden (aged 16-79 years) regarding victimization, fear of crime and public confidence in the justice system”. The 2015 report is based on approximately 11,900 responses which represents a 60% response rate (responses are obtained via telephone interviews, postal questionnaires and online questionnaires)
Looking at the police-reported crimes figures first, I found that in 2015, reported sexual assaults actually fell back from their 2014 high to the levels reported in 2013. So this sounds like good news.
The Swedish Crime Survey, however, tells a different story entirely. There is a 2-year lag between the period being surveyed and the publication of the report, so the most recent survey, published in February 2017 and conducted during 2016, actually relates to participants’ experiences during 2015. According to the survey,
In 2015, 1.7 per cent of persons stated that they had been exposed to a sexual offence. This is an increase as compared with 2014, when 1.0 per cent stated that they had been exposed. Sexual offences remained at a relatively stable level for the period 2005-2012 with approximately 1 per cent exposure, and an increase occurred thereafter.
This amounts to a 70% increase in reports of sexual offences over the space of a year, which initially appears quite alarming. The problem, of course, is the distinction between “violent sexual assault” (as recorded by the UNODC) and “sexual offences” (which would include but presumably not be limited to sexual assault). Sexual offences could include online or verbal harassment or any other number of non-violent offences, so this does not mean there has been a 70% increase in violent sexual assault. It is, however, still a worrying increase assuming no change in categorisation between 2014 and 2015, as to what is considered a sexual offence. It is also worth pointing out that according to the Swedish Crime Survey, only 9% of sexual offences are reported to the police, though no explanation or theory is provided as to why this figure is so low (it could be that the majority of the offences are relatively minor, but on the other hand it could indicate a level of mistrust in the police taking offences seriously or a lack of faith in the likelihood of a successful prosecution).
Trying to find comparative figures for the same period for the UK, I referred to the statistical bulletin published by the Office of National Statistics. According to the ONS, the UK saw a 37% increase in police-recorded sexual offences in 2015 compared to 2014 and a further 21% increase in 2016. The ONS report points out, however:
The increase in police recorded sexual offences should be seen in the context of a number of high profile reports and inquiries which are thought to have resulted in police forces reviewing and improving their recording practices.
In conclusion then:
- Sweden has seen an increase of 67% in police-recorded sexual assault between 2003 and 2015.
- Most of this increase, however, occurred between 2003 and 2010, likely due to changes in the way crimes were recorded during this period. The increase between 2010 and 2015 is only around 1%
- Swedish crime survey results (which include sexual offences that are not reported to the police) tell a different story, showing stable levels from 2005 to 2012, at around 1% of the population being victims of a sexual offence, increasing thereafter to the 2015 level of 1.7% of the population (it actually increased to 1.3% in 2013, falling back to 1% in 2014 before increasing again to 1.7% in 2015)
- In England and Wales, by comparison, levels of police-recorded sexual assault increased by 68% between 2010 and 2015. As in Sweden for the 2003-2010 period, this is being attributed to improvements in recording practices between 2010 and 2015 in England and Wales, as well as a number of historic sexual offences being recorded during this period in the wake of the Jimmy Saville scandal.
- The Crime Survey for England and Wales, by comparison, shows relatively stable levels, of around 2% of the population having been a victim of a sexual assault, with some minor fluctuations year on year.
Allowing for increases in police-recorded sexual assaults that are attributed to improvements in police reporting, the statistic that stands out is the 70% increase in reports of sexual offences within the Swedish crime survey from 2012 to 2015. Until the 2016 data is published it is impossible to hypothesise over whether the sharp increase in 2015 has since been followed by a correction, or whether the level is still the same or continuing to increase.
So is Sweden the rape capital of the world? Given the fact that many countries don’t have proper sexual violence laws, or enforcement of those laws, no, it’s clearly not fair to make that statement based purely on the higher number of recorded incidents in Sweden than elsewhere. But is there a problem with increased levels of sexual violence in Sweden? Given the Swedish Crime Survey’s claim that only 9% of sexual offences are reported to the police, and the not-yet-available survey results for 2016, I will continue to keep an open mind.