Tuesday, August 02, 2005

Blood lines

It is a fact that the majority of the suicide bombers who attacked London where Middle-Eastern or North-African looking.

Now many might intuitively think this is an excuse for targeted, racially profiled, stop and search. After all, most people's common sense will tell them, that if the police are searching for football hooligans, it makes no sense to stop and search an elderly Japanese woman. Ditto, Islamist terrorists.

It may be intuitive, but this thinking is flawed. All it ends up doing, is advertising to the terrorists what your (institutionalised) blind spot is. How long will it be before they draft into service someone who doesn't fit the profile?

Funily enough, the "statistical" approach is quite common. Take the ban on gay men donating blood.

The argument against not aloowing gay men to donate blood is based soley on the fact that gay men are more statistically likely to be carrying the HIV virus - but not exculsively so, because there are many straight men (and women) who carry the virus. Therefore, it makes sense to check all blood that is donated - because a statistical approach won't be foolproof.

It is also a fact that statistically, a Middle-Eastern or North African man is more likely to be a suicide bomber, but not exclusively so, merely "statistically". Because it is not exclusisely so, it makes sense to profile on behaviour, body language, etc, rather than on race or religion or gender.

In both cases the stakes are high: if blood is evaluated on statitics, it won't be safe - that is why every pint donated needs to be checked (and I believe is - so why the ban? Is this not a tacit admission that the 'statistical' argument is so flawed as to be useless - and only prejudice remains?).

Similarly, to be blinkered by statistical profiling in finding bombers will expose the same deficiencies in that approach.

POSTSCRIPT
Just found this story from Australia on gay.com. Must be zeitgeist.
"Gay blood donor ban criticised"
http://uk.gay.com/headlines/8850

0 Comments:

Post a Comment

<< Home