Fun with numbers: Racist crimes

Old Holborn posted the following racist crime stat's on Sunday (click to enlarge):
The conclusion being that there were thirty times as many white victims of racial violence as you would expect, which set off the inevitable huffing and puffing.

As you will see if you read the comments, I walked right into it. I pointed out that on the basis of those figures, a black person is four times as likely to become a victim of a racist crime than a white person. It was pointed out to me that the more important point was that a black person was thirty times as likely to be a perpetrator of a racist crime than a white person, which is quite correct (in fact, subject to certain assumptions, the figures might be five times as likely to be a victim and sixty times as likely to be a perpetrator) at which stage I withdrew my original comment.

Hmm. I cannot track down the statistics referred to, not after an hour's worth of googling, and the best 'official' figures I can find are that "Less than one per cent of the White population had been victims of racially motivated crimes compared with two per cent of people from Asian, Black and Chinese and Other ethnic backgrounds" (from page 11, Statistics on Race and the Criminal Justice System – 2006) or that "people from minority ethnic groups are ten times more likely to be a victim of a racist incident than people from white ethnic groups" (from page 2, Better Housing Briefing, Racist harassment and housing services).

That's one heck of a discrepancy (twice as likely or ten times as likely?) so let's stick with five times as likely as a mid-figure. Assuming, in round figures, that there are 57 million white people and 4 million non-white people in this country, and there were 179,000 "race and faith hate crimes" in the year concerned (2007 Hate Crime Survey - Statistics and Trends: A Review of Select Countries), there must have been 132,000 white victims and 47,000 non-white victims.

We can't even begin to speculate on the number of racist attacks perpetrated on one minority group by other minority groups, but it must be pretty safe to say that outside of Northern Ireland, racist crimes against whites will be perpetrated by non-whites. Ergo, 4 million non-white people are responsible for 132,000 racist crimes against white people (an average of one crime per 30 non-white people), and even if we make the heroic assumption that all racist crimes against non-whites are perpetrated by whites, that is still only an average of one crime per 1,213 white people*.

So I have a nasty feeling that the original "thirty times more likely" figure is in fact probably correct.

Just sayin', is all.

* Using the same methodology and the same total number of racist crimes, if whites and non-whites had the same propensity to commit racial crimes, non-whites would actually be two hundred times as likely to become a victim; and conversely, if non-whites were two hundred times as likely to perpetrate racial crimes, the chance of a member of either group becoming a victim thereof would the same.