“Flying while Muslim” is the new “driving while black”, according to air travellers who believe they are being targeted for extra security measures on the basis of racial and religious profiling [New Scientist]. The Transportation Security Administration is mum on whether they use racial profiling in deciding who to pull out of line at airport security, and also won’t give details as to whether it scrutinizes air travelers’ behavior (like when and how they bought their tickets, or whether they have checked luggage). But a new mathematical study suggests that any such profiling is not the most effective way to find a terrorist lurking in a crowd of ordinary people.
At first glance, the profiling approach seems logical, despite many people’s moral objections. If all previous acts of politically motivated terrorism have been committed by a particular nationality, then doesn’t it make sense to focus searches on those groups? Not necessarily, says William Press of the University of Texas at Austin. Do the maths and you discover that a simple-minded application of these actuarial methods is worthless: all you end up doing is repeatedly picking out the same innocent people [Nature News].
In the study, published in the Proceedings of the National Academy of Sciences, Press modeled situations in which members of a profiled group are, for example, 100 times more likely to be a terrorist than a typical traveller. You might think that the best approach would be to make it 100 times more likely that these people are selected for extra security checks. But in fact, random screening turns out to be just as effective. The problem is that too much time is spent repeatedly screening members of the profiled group who are not actually terrorists, Press explains [New Scientist].
Press did find a slight benefit to profiling in a weakened form. Mathematically speaking, it would be optimal to screen individuals in proportion to the square root of their presumed probability of malfeasance [Scientific American]. Using this “square-root-biased sampling,” a member of a group that is considered 100 times more likely to contain terrorists would be 10 times more likely to get extra screening. This would distribute resources more evenly, effectively casting a wider net and catching more would-be terrorists.
But Press says real-world terrorist profiles are notoriously hard to construct, and it’s extremely unlikely that any group–be it an ethnic group or people who pay cash for their tickets–could be considered 100 times more likely to contain terrorists. A more realistic probability factor might be 10, and taking the square root of that means that members of the group would be only about 3 times more likely to receive extra screening. Press says that at that point the profiling is so watered down and the benefit is so small that this should “reopen the moral and ethical questions of whether the profiling is worth the social cost at all,” he says. “Personally, I would say that it is not” [Scientific American].
DISCOVER: The No-Touch Pat-Down describes the “air puffer”
DISCOVER: Future Tech: facial recognition systems in airports
Discoblog: Co-Ed Naked Airport Security: X Ray Scanners Strip Search Passengers
Image: flickr / nedrichards