What an irony. You would think that the publishing of hate crime statistics would be an annual opportunity to raise the profile of hate crime and also to renew efforts to prevent it. But no, the forums are alive with spurious interpretations of what it all means and even attempts to show that Somalis are in fact more racist than Finns. Can it be true?
One effect of hate crime statistics being published in Finland is that it brings up once again the unwelcome question of whether Finns are more racist than other nations. This isn’t my question, by the way, but it is one that Finns tend to dwell on, as if there were an acceptable level of racism that a country is allowed to have!
What has also been happening is that international comparisons are made for the statistics within Finland, so that means looking at the different nationalities living in Finland, and providing a league table of the racists among them.
This isn’t just something you will find in the hate forums though. On page 58 of the 2011 Police report (recently published by the Police College of Finland and the Ministry of Interior’s Police Department), the researchers themselves provide their own league table (Table 12) detailing racist crimes by foreigners per 1000 head of population, with Somalis top of the list on 10 per 1000 head of population, followed by Iraqis on 8. A quick calculation gives an equivalent rate for Finns of 0.1 per 1000 head of population. Wow, aren’t the Finns so much better than those dirty foreigners? Ugh, well, no actually! Quite the opposite! But you’ll have to bear with me, I’m afraid, because you see, they haven’t bothered to inform the reader just how useless and distorting this kind of statistic actually is.
Several figures for rate per capita are possible here. For example, we could look at those racist crimes that do not involve provocation by another racist. We could also look at the age group 15-34, as this is where the biggest problems lie (66% of racist crimes) and it’s also the segment of the population that is most different in proportions for the different nationalities (see below). Narrowing the focus doesn’t change any of the fundamentals in the argument, by the way.
These simple adjustments give us the figures of 0.028 attacks per capita by Finns aged 15-34, and 0.14 attacks per capita by Somalis of the same age.
It still looks grim for the Somalis, doesn’t it? But calculations made per capita are horribly flawed.
They ignore the fact that many Finns will seldom or never encounter a black immigrant, while the same is not true for black immigrants meeting white Finns.
Let’s put it plain and simple. If 1000 in a 10,000 Finns actually get to meet a Somali (or any other visible foreigner), and one of these Finns is a racist who decides to assault the foreigner, then it’s all too easy to claim that only one in 10,000 are racist because only one in 10,000 assaults showed up ‘per capita’, when in reality, it was one in 1000 Finns who got to encounter a foreigner and were racist. The opposite effect works for Somalis, who will likely encounter people of a different race many times daily.
Let’s show the maths
Two population groups, A and B, where A is a minority and B is a majority. Let’s set the rate of racism at 5% for both Group A and Group B, so we would therefore expect the per capita crimes to be the same. If they are not, then something is wrong with using per capita rates to define rates of racism.
Two key assumption, a racist assault can only be committed by someone who encounters someone else of a different race or ethnicity. One kind of racist is someone who has great difficulty in encountering someone of another race.
Group A – minority
N = 100 (total population of minority)
n = 100 (number of population encountering people of another race)
C = number of crimes per year = 5 (i.e. 5% of n)
per capita rate = 0.05 (based on C/N)
Group B – majority
N = 2500
n = 250 (number of population encountering people of another race)
C = number of crimes per year = 12.5 (5% of n)
(1) per capita rate = 0.005 (based on C/N, i.e. not taking account of n)
(2) per capita rate = 0.05 (based on C/n, i.e. taking account of n)
The first calculation for the per capita rate of racist crime in the majority population (1) is 10 times greater that for the minority than the majority population, even though WE set the rate at EXACTLY THE SAME LEVEL, i.e. 5%, so as to control for it. In equation (2), the rate of encounter has been factored in and it delivers the correct rate of crime per capita.
Conclusion, the rate of encounter can have an enormous effect on statistics of rates of racist crime per capita.
if we used calculation (1), we would be throwing up our arms and saying that the Finns are getting attacked so much more and that the Somalis are ten times more racist, because when you do the maths, that’s what you get (exactly like the Finnish Police have suggested we use the statistics).
Finnish racists when they exist simply have less opportunity to express their racism and so to appear in the statistics. Using the ‘whole population’ therefore waters down the ‘per encounter’ rate and hugely underestimates the rate of racism of the majority population.
To discover the true rate of racist crime, we must adjust for the fact that minorities are by definition rare and encounters with them are likewise rare for the majority population, while for the minority, encounters with Finns are commonplace. We must take account of rates of encounter to arrive at anything like a comparable per capita rate. Hate crimes are available in part to estimate rates of racism. It should be absolutely clear where the per capita figures are horrendously flawed in estimating rates of racism. Common sense, folks.
In other words, adjusting for the Somali minority in the age group 15¬-34 would give us a rate that is 227 times smaller than it currently is and a long way behind the rate for Finns.
Finns = 0.012 attacks per interracial encounter
Somalis = 0.000083 attacks per interracial encounter
Of course this isn’t the end of the story. There are numerous other factors that can also impact disproportionately on the statistics, including the higher percentage of young adults as a percentage in the age group aged 15-35, who are most likely to be involved in street assault. For Finns, they constitute about 23% of the population, while for Somalis, they constitute 40%. This would artificially inflate the Somali figure even further.
The idea of league tables per capita that compares minority populations or by implication the majority population is inherently flawed and is in my view an extremely cynical abuse of statistics where there is no explanation of the drawbacks. The work of a racist researcher? Or perhaps a totally incompetent one. Either way, it’s not good enough.
These issues are too important to think that people haven’t stopped to ask themselves what are the underlying assumptions in these statistics and how are they likely to be used, or to have actually monitored how the statistics are being used in the public debate. One look at the comments here or on some of the hate forums reveals just how eager racists are to make use of the police report to disparage immigrants.
Why do we have a league table, if not to make foreigners look bad and to make Finns look better than they are or to hide the true extent of racism in Finland? The idea of a league table is abhorrent and extremely misleading. It’s worse than that, it’s lying.
What adds to the injury is that people rely on these statistics to create profiles of particular national groups as being much more racist than they actually are, and much more racist than Finns.
So, hate crime statistics that are presented in such a way that they actually perpetuate hate crime!
Surely not, Finland, surely not!!!! Tell me I’m wrong!
PS. this article has been edited for the sake of clarity.