Friday, January 26, 2007

Rape Conviction Rates.

In the light of the rejection by judges of the government's proposals to change the law regarding rape I have been thinking about the oft cited low conviction rates for rape allegations. The proposals themselves are as ever with this government extremely silly and involve things like redefining rape to include things like having sex whilst drunk. Six percent does seem like a remarkably low proportion of convictions to allegations but is it necessarily so or is it merely poor statisitical reasoning to assume it is so?

Suppose a test for a disease is 99% accurate, ie if 100 people with the disease are tested 99 will be shown by the test to have it and if 100 without the disease are tested then 99 will be shown to be clear. However if the disease affects only 1 in 10000 people then how likely is someone who is diagnosed with the disease to actually have it? The answer is of course not 99%, in fact because 1% of the results are wrong and the disease is rare the person who tests positive is around 100 times more likely to be negative for the disease.

So at first a conviction rate of 6% appears extremely low, but because far more women are not raped during a given period than are raped this is not necessarily so. Suppose 500 women are raped in a month and one in 5 report the crime. Suppose also that 10 000 000 women are not raped, but for whatever reason one in 100 000 decide to make a bogus allegation to the police. In this situation half of the rape allegations received by the police would be bogus, despite the fact that only 0.001% of women are anything other than completely honest about the situation.

Now these figures are obviously random, but no more so than the various ministers and columnists who assume that a 6% conviction rate means that the other 94% are examples of the system letting rapists get off. It certainly does not justify newspaper columnists screaming "Any man reading this article can go out tonight, pick out a woman, force her into sex, and be almost entirely certain he will walk free"

There is a strand of thought amoung political extremists that no one ever lies about being raped but recent high profile cases like the Duke Lacrosse case or the Warren Blackwell case or Timothy Gee demonstrate that this is simply preposterous, people lie about all sorts of things. This is why the disturbingly widespread belief that the normal standards required for a fair trial should be dismissed in rape cases is so very wrong.

1 comment:

Anonymous said...

Good article, nice and logical. The prevalence of false rape allegations (I believe) is due to (as you say) the definition of rape being expanded to include all sorts of things, and the protection of all female accusers of rape, whether it actually happened, or the woman lied. As in the Blackwell case.

A situation exists where these women can break the law without punishment. False accusers should be able to be punished. A man can be put on the sex offenders register based on nothing but a woman's word. Therefore a false accuser should be put on the register, or create a register of 'false accusers' or somesuch list.