Well, racial profiling in 'war on terror' will only make terrorists' job easier. What better place to hide a weapon than unsuspecting grandma's huge handbag? In pre-1948 Palestine both Jews and Arabs were choosing light-skinned guys to attack British...

Edit: again not reading the thread to the end before responding redundantly....

Well, thanks, though, for reminding me that I forgot to respond to Jeff's response.

My thought in using the "grandma" example was not to say that some element of randomness is bad, but I think that complete randomness (as practiced by early post-9//11 secondary screeners) is/was bad and pretty hilarious. I *do* have a hard time worrying that Islamist terrorists will recruit 87 and 85 year-old Sven and Shirley Jensen from Cedar Rapids, Iowa to do their dirty work. Not impossible, but likely? Perhaps as the Dept of Homeland Security's brilliant threat analyses identify an uptick in threats from the domestic Timothy MvVeighs, then Sven and Shirley should get a harder look. On the flip side, relying lazily on profiling to just pull all the "Middle Eastern-looking men" out of line for screening is a setup for the kind of unhappy surprise you suggest.

Some element of randomness increases the likelihood that flaws in other screening methods will be identified, but I would say that complete randomness is not effective without cost-prohibitive screening rates.

If the reputation of El Al is to be believed, they have a successful (the most successful?), albeit expensive, system that is multi-factorial.

To your point of working around "profiling" preconceptions with atypical bombers, I would, in a brief tangent, recommend the film The Battle of Algiers to anyone who has never seen it. Great movie.