29

The US think-tank, the Gatestone Institute - along with many conservative web-sites, has claimed that rapes in Sweden have gone up because of the influx of refugees from the Middle East.

Forty years after the Swedish parliament unanimously decided to change the formerly homogenous Sweden into a multicultural country, violent crime has increased by 300% and rapes by 1,472%. Sweden is now number two on the list of rape countries, surpassed only by Lesotho in Southern Africa.

[...]

Over the past 10-15 years, immigrants have mainly come from Muslim countries such as Iraq, Syria and Somalia. Might this mass influx explain Sweden's rape explosion? It is difficult to give a precise answer, because Swedish law forbids registration based on people's ancestry or religion. One possible explanation is that, on average, people from the Middle East have a vastly different view of women and sex than Scandinavians have. And despite the attempts by the Swedish establishment to convince the population that everyone setting foot on Swedish soil becomes exactly like those who have lived here for dozens of generations, facts point in an altogether different direction.

Are these correlations accurate? Is the proposed causality plausible?

inund8
  • 591
  • 4
  • 20
  • 1
    [Related question](http://skeptics.stackexchange.com/questions/18700/did-muslim-immigrants-rape-over-300-swedish-children-in-seven-months-of-2013) about child rape. – Oddthinking Dec 18 '15 at 01:34
  • 17
    I can't address the violent crime numbers but the rape numbers have a big problem: They expanded the range of actions they class as "rape", thus making it impossible to compare old numbers with new. (Remember the rape allegation against Assange? The alleged act was not wearing the agreed-upon condom.) – Loren Pechtel Dec 18 '15 at 02:48
  • With a reference that would be the basis of a good answer. – DJClayworth Dec 18 '15 at 04:34
  • 1
    Useful article: http://www.bbc.com/news/magazine-19592372 – DJClayworth Dec 18 '15 at 04:41
  • 4
    @Loren http://skeptics.stackexchange.com/questions/8498/is-the-julian-assange-sexual-assault-charge-merely-alleging-condom-non-use – Andrew Grimm Dec 18 '15 at 08:01
  • 7
    Definitions have changed. See also [Wikipedia](https://en.wikipedia.org/wiki/Rape_in_Sweden). International comparisons of rape statistics are effectively impossible. – gerrit Dec 18 '15 at 11:30
  • 4
    Sounds like a false correlation/causation assertion to me, even if the figures are correct (and I don't have any evidence to say that they are). You could argue that including the Eastern European countries in the Eurovision Song Contest was equally to blame, or switching from analogue to digital communication, or anything that has happened in the time period between the two samples. – GordonM Dec 18 '15 at 12:46
  • 1
    @markbiernacki Plausibility is a subjective opinion and should IMHO not be queried here. The Gatestone article you're linking to does not try to hide that they are also just speculating: "Might this ...? It is difficult to give a precise answer, ..." The numbers they are quoting (mostly with links to the source) seem legit. They are also right that limitations in public Swedish crime statistics makes it impossible to prove a relation between immigration from the Middle East and the rise in the number of reported rapes. – Tor-Einar Jarnbjo Dec 19 '15 at 13:00
  • 1
    Not exactly the same question, but in my answer, I try to point out some of the deficiencies in Swedish crime statistics: http://skeptics.stackexchange.com/questions/18700/did-muslim-immigrants-rape-over-300-swedish-children-in-seven-months-of-2013/18740#18740 – Tor-Einar Jarnbjo Dec 19 '15 at 13:04
  • 6
    The hinted causation is contradictory with the article itself. If Sweden was second country in the World for rape (an meaningless claim without a world-wide shared definition of "rape"), it would be doing worse than Middle East countries. If anything, immigration from these countries of men with "vastly different view of women and sex than Scandinavians" should then decrease the trend for rape in Sweden, not increase it... – Evargalo Nov 09 '17 at 12:45
  • 1
    Another point that may be of interest to readers is that while right-wing propaganda seems to imply that immigrants are threatening Swedish women, the actual victim profile is not discussed - according to the [Globe and Mail](https://www.theglobeandmail.com/opinion/swedens-rape-crisis-isnt-what-it-seems/article30019623/) article: "... almost all the victims of these crimes – especially sex crimes – are also foreign-born." – Arnon Weinberg Mar 24 '19 at 06:01

1 Answers1

46

No, there is not even a good correlation between actual rape rates and immigration.

The figure that Gatestone Institute cites is the rate of reported rapes. This has indeed increased, but the reason has nothing to do with a supposed increase in actual rapes.

This is because crime victim surveys[1] have shown that the actual rates of rape have remained relatively unchanged from 2005 (when measurements started) to 2014 (latest year for which summary statistics is available). So we know that actual rates of rape are more or less constant during that period and violent crime has decreased overall for decades (according to The Better Angels of Our Nature: Why Violence Has Declined by Steven Pinker).

So why have the rate of reported rapes increased? There are a few reasons:

  • Expansion of rape definition in several steps: actions that are comparable to rape (1998)[4], helpless state (2005)[5] and particularly vulnerable situation (2013) [6][2].

  • Increased tendency to report crime: the observed tendency to report crime, as measured by number of rape reports / number of rapes in crime victim surveys has doubled between 2005 (10%) and 2010 (20%) [3].

  • Changes in how police handle rape reports: Swedish police makes one police report per rape, so if a person has been raped 30 times in a relationship, it will count as 30 separate rape crimes and they have made a conscious effort to file all sex crimes that could be rape as rape even though they might end up as being another sex crime or no crime at all [2].

So there is not even a correlation to speak of. Sweden has had lots of immigration during 2005-2014, yet more or less constant actual rates of rape.

Anti-immigration activists often retort by claiming that immigrants are overrepresented in crime statistics and while true, the observed overrepresentation is smaller for immigrants (2.5x) than for men (3.5x) and people who are unemployed, on welfare or without high school education (5x-6x), so they are blowing it way out of proportion [7].

References:

All references (except the last one) come from the Swedish Council for Crime Prevention (Brottsförebyggande Rådet, BRÅ), which is a Swedish governmental agency that collects statistics and writes reports about crime. The last reference is from the National Centre for Knowledge on Men's Violence Against Women, University of Uppsala.

Unfortunately, they are almost all in Swedish (since they are primary sources), but BRÅ is a recognized authority on crime statistics in Sweden, even among the anti-immigration activists.

[1] https://web.archive.org/web/20170228040947/https://www.bra.se/bra/brott-och-statistik/statistik/utsatthet-for-brott/ntu.html (red line is sex crimes, yearly reports contain information about rapes specifically)

[2] https://www.bra.se/bra/bra-in-english/home/news-from-bra/archive/news/2011-01-18-how-common-is-rape-in-sweden-compared-to-other-european-countries.html

[3] https://www.bra.se/bra/nytt-fran-bra/arkiv/press/2012-11-06-allt-fler-polisanmaler-nar-de-utsatts-for-brott.html

[4] https://www.bra.se/download/18.cba82f7130f475a2f1800012829/2005_07_valdtakt_kartlaggning.pdf (p. 16)

[5] https://www.bra.se/download/18.744c0a913040e4033180001042/2011_6_polisanmalda_valdtkter_barn.pdf (pp. 9-10)

[6] http://www.nck.uu.se/Kunskapscentrum/Kunskapsbanken/amnen/Sexuellt_vald/Sexualbrottslagstiftningen/

[7] https://www.bra.se/download/18.cba82f7130f475a2f1800012697/2005_17_brottslighet_bland_personer_fodda_sverige_och_utlandet.pdf (p. 35 )

gerrit
  • 17,636
  • 17
  • 84
  • 137
EmilKarlsson
  • 698
  • 7
  • 5
  • 12
    You are misquoting the public statistics at several points. The first survey only *estimates* the number of incidents and does not provide any *actual* values. The relatively constant rate you're referring to is the number of sex related crimes in total and not the number of rapes. The estimated number of rapes has changed significantly and risen from 18,000 in 2009 to 70,000 in 2013. The report points out that these numbers are potentially very inaccurate, since they are based on the victims perception of the situation and not on the legal definition of 'rape'. – Tor-Einar Jarnbjo Dec 20 '15 at 14:12
  • 9
    Furthermore, the overrepresentation of immigrants in the category 'rape or rape attempt' is 5.0 and not 2.5 as you're quoting. The number can be found on page 43 of the report you are referring to. – Tor-Einar Jarnbjo Dec 20 '15 at 14:16
  • 13
    @Tor-EinarJarnbjo 1. Of course it estimates the number of cases. That is what all studies do. No study of crime has the entire population as the sample. You are simply being facetious. I use the term actual in contrast to reported to refer to crime victim studies instead of reported figures. – EmilKarlsson Dec 20 '15 at 14:31
  • 5
    @Tor-EinarJarnbjo 2. I clearly state in ref [1] that the red line is all sex crimes and refer to individual yearly reports. The figures you cite are cherry-picked and during the entire period for which we have data, the actual ("estimated") rape rates are fairly constant as a proportion of total population. Variation year to year is expected and they ask additional questions to tease out what kind of sex crime / what kind of rape. – EmilKarlsson Dec 20 '15 at 14:33
  • 12
    @Tor-EinarJarnbjo 3. The overrepresentation I cited is for all crimes, which I specifically wrote. I never talked about overrepresentation for rape or attempted rape. Of course your objection does not hold any water, since men are even more overrepresented in sex crimes / rapes / attempted rapes than crime in general. The general point here is that anti-immigration activists freak out over immigrant crime (including rape / attempted rape), but hardly bat an eye when it comes to these other groups, which exposes a key inconsistency / contradiction in their position. – EmilKarlsson Dec 20 '15 at 14:35
  • 3
    You write in your answer that the survey 'has shown that the actual rates of rape have remained relatively unchanged from 2005'. The survey does not show this and you back your statement by pointing to a diagram showing a different value (number of sex crimes in total). The figures I cite are not cherry-picked, but the estimated number of rapes stated in the yearly reports, which you even mention yourself without quoting the values and otherwise seem to ignore. – Tor-Einar Jarnbjo Dec 20 '15 at 14:58
  • 1
    For the overrepresentation of immigrants, you also pick a more generic figure (for all crimes, instead of for 'rape and rape attempt'). Since the question is about the development of the rape rate over a 40 year period, going over only a 10 year period and using far too generic figures as an argument against Gatestone's assumpotions (although the statistics provide more detailed figures directly contradicting your argumentation) is not particularly helpful. – Tor-Einar Jarnbjo Dec 20 '15 at 15:02
  • 4
    The survey does show that, since the individual yearly reports estimate the number of rapes and barring year-to-year variation, there is a more or less constant trend in terms of % of population. Additional data from the Pinker book shows that there is a long-term decline. Your values are cherry-picked because you picked two years instead of looking at all years. You also have no response to the problem with year-to-year variability. The question is not about a 40 year period, but a 10-15 year period: "Over the past 10-15 years [...] Might this mass influx explain Sweden's rape explosion?". – EmilKarlsson Dec 20 '15 at 15:40
  • 5
    The most increase in immigration has been since 2005, so if immigration caused massive increase in actual rapes, it would show up here. But it does not. Moving the goal posts to 40 years when you have been disproved is intellectually dishonest. – EmilKarlsson Dec 20 '15 at 15:42
  • 1
    Even if you look at the overrepresentation for rapes and attempted rapes, the argument still holds since men are more overrepresented in rapes / attempted rapes than all crime combined. You have not addressed this argument whatsoever. – EmilKarlsson Dec 20 '15 at 15:44
  • 1
    As additional evidence that you cherry-picked your figures, 2013 has the highest estimate of rapes for the entire time period (likely due to year-to-year variability). – EmilKarlsson Dec 20 '15 at 15:46
  • The number of estimated rapes for the years 2007 to 2013 are (in thousands): 25, 25, 18, 18, 29, 36, 70. I can't find the yearly reports from 2005 and 2006 and the reason why I used 2013 as the last year is simply that the preliminary report for 2014 does not provide the number of rapes, but combines rape with the lesser offense 'sexuell tvång* and counts 96.000 cases. I have not addressed your comparison with men regarding the overrepresentation since I am not sure why that should be relevant. – Tor-Einar Jarnbjo Dec 20 '15 at 15:59
  • The relevant number is the overrepresentation of immigrants in the rape statistics and not if you can find other population groups, which also are more likely to conduct rape. – Tor-Einar Jarnbjo Dec 20 '15 at 16:04
  • So you acknowledge now that your selection of 2013 was cherry-picking, since it looks like it is due to year-to-year variability taking into account the decreased observed in 2014 compared with 2013 in sex crimes? Good. Provided 2013 is an outlier, looking at the other figures and taking into account population increases over time, the estimated number of rapes as a proportion of the population is either more or less the same (allowing for year-to-year variability) or declining. – EmilKarlsson Dec 20 '15 at 16:50
  • 4
    Yes, looking at other groups are very relevant, since you must evaluate the estimated statistics in the scientific context. We want to know if a 2.5 (or 5.0) overrepresentation is a tiny, small, moderate, large or huge overrepresentation. When compared with other, larger groups, such as men, poor, people with low education and low income, the immigration overrepresentation, whether all crimes or just rapes/attempted rapes are not very practically significant. – EmilKarlsson Dec 20 '15 at 16:52
  • 3
    It also highlights a substantial logical inconsistency in the anti-immigration position in that they freak out over immigration overrepresentation, but do not bat an eye at the considerably higher overrepresentation for other groups. In other words, their position is inconsistent. – EmilKarlsson Dec 20 '15 at 16:53
  • I am not sure if it makes sense to repeat myself, but you don't seem to read what I am writing. I have absolutely no idea where you take it from that I agree with you that using the data from 2013 as the last comparison value was cherry picking. I just explained to you that I used the number from 2013 because it is the last data availabe, since the 2014 report uses a different method of counting. – Tor-Einar Jarnbjo Dec 20 '15 at 17:00
  • 4
    If you look at the figures you posted: 25, 25, 18, 18, 29, 36, 70, it is clear that the 2013 value (70) is an outlier, especially since this is the case for sex crime data overall, since 2014 show a decline on percentage data. When you pick outliers to demonstrate a point, that is cherry-picking. It simply does not matter that 2013 is the latest data available. I am not claiming that you did this intentionally to deceive, I am simply pointing out that it looks like an outlier and thus an inappropriate year to pick. – EmilKarlsson Dec 20 '15 at 17:08
  • 1
    I think we have both outlined our positions, objections and rebuttals in sufficient detail for readers to consider, so unless there is any qualitatively new argument being presented from your end, I am more or less satisfied with this exchange. – EmilKarlsson Dec 20 '15 at 17:09
  • 15
    @Tor-EinarJarnbjo Consider posting your own answer. – Jonathon Dec 21 '15 at 02:25
  • 4
    @EmilKarlsson That makes no sense. The entire point of this questions is, have rape rates exploded. AKA, we are looking if the last year(s) are outliers (extremely high outliers). You put forth some other possible causes, but you cannot just lie about the data, throw out all "outliers" and tell people that "no, rape rates have remained consistent and unchanged" – Jonathon Dec 21 '15 at 02:30
  • Comments are not for extended discussion; this conversation has been [moved to chat](http://chat.stackexchange.com/rooms/33261/discussion-on-answer-by-emilkarlsson-is-sweden-experiencing-a-rape-epidemic). – Larian LeQuella Dec 21 '15 at 04:06
  • 1
    @JonathonWisnoski Pointing out that 2013 is an outlier, with reference to data from 2005-2014 that clearly indicate that 2013 is an outlier is not "lying about the data". Even if it was not an outlier, it still makes no sense from the immigrants-cause-rape-explosion perspective: they only did it for 2013, but not earlier or later years? Makes no sense whatsoever. – EmilKarlsson Dec 21 '15 at 14:00
  • 4
    @EmilKarlsson Except that is not even what you did. You state "have remained relatively unchanged" full stop. You never mention outliers. Which is why, latter when you talk about why the rate is increasing, but it is not the immigrants fault, it makes no sense. There is nothing stable about that data. We see it do down to 18, then start spiking upward, 60% in 2011, 30% in 2012 (it has already doubled), and finally it doubles again in 2013. Even if we remove 2013, we still have a spike the goes to double the local minimum of 18, which was the rate 2 AND three years prior. – Jonathon Dec 21 '15 at 16:27
  • 3
    So are we going to consider three years of data to be outliers, because some preliminary, data for 2014 and 2015 seem to suggest it might be back to normal? – Jonathon Dec 21 '15 at 16:29
  • 1
    @JonathonWisnoski I would have posted my own answer if the available statistics had provided any data, making it possible to answer the question. I've already pointed out in comments to the question that the question probably cannot be answered. – Tor-Einar Jarnbjo Dec 21 '15 at 23:49
  • @EmilKarlsson If immigrants are overrepresented in rape-like crimes by the same factor that men are overrepresented in crime in general, that points to it being a huge overrepresentation, right? – sgf Jul 15 '17 at 12:55
  • 2
    @sgf It depends: if men are also overrepresented among immigrants, not at all... – Evargalo Nov 09 '17 at 12:59
  • Actually, your answer do not refute far-right wings claims. Because your definitions of immigrants are not exactly the same. Their claim include actuals immigrants and descendants of previous immigrants. The latter may prove more trouble than the former. I have been quite surprised to see the comtempt some of my close immigrants friends (muslims themselves) hold against descendant of muslim immigrants. It would be interesting to see serious studies about factors like population heterogeneity, density of population, dating habits, real income, ... – xrorox Feb 26 '18 at 16:36
  • Some other factors that may be interestings would be ratio between males and females, family structure, ect ... Alas, it would mean to drop the politically correct, in order to measure which factors are relevant directly or indirectly. But anyway someone is going to be offended. – xrorox Feb 26 '18 at 16:39