The math of security screeningRandom checks as effective as terrorist profiling
Profiling is a waste of time and resources in security screening; the problem is that too much time is spent repeatedly screening members of the profiled group who are not actually terrorists,
African Americans have an expression for it: DWB, for “driving while black.” This phrase captures the fact — statistically proven fact — that police officers across the United States stop black drivers, for real or imagined traffic violations, much more often than they do white drivers. There is now another expression — FWM, for ”Flying while Muslim.” This expression describes air travelers who believe they are being targeted for extra security measures on the basis of racial and religious profiling.
Not only is profiling discriminatory, but it is also inefficient, according to a new analysis by William Press of the University of Texas at Austin. The New York Times’s Sandra Blakeslee writes that his argument applies not only to crude racial profiling, but also to profiles that flag up risky individuals on the basis of how they bought their tickets, whether they have checked luggage, whether they have purchased one-way flights, and so on. Press has modeled situations in which members of a profiled group are, for example, 100 times more likely to be a terrorist than a typical traveler. You would think that the best approach would be to make it 100 times more likely that these people are selected for extra security checks. In fact, though, 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. For the same reason, a strategy in which members of the profiled group are always pulled aside performs more poorly than random screening, diverting resources from members of the “low risk” majority who may still be terrorists.
A better approach, Press says, would be to take the square root of the extra risk posed by the profiled group, and use that number to define the likelihood that someone from the group will be pulled out of line. Under this scheme, members of a profiled group who are 100 more likely to be a terrorist than the typical traveler would be 10 times more likely to be given secondary checks.
Given the rarity of actual terrorists, however, this still may not be much better than random screening. “We probably shouldn’t be doing profiling at all,” Press suggests, once you weigh up the debatable benefits against concerns about discrimination.
In any case, quantifying the risk posed by a profiled group may be extremely difficult. “Fortunately, terrorists are few and far between,” says Barry Steinhardt of the American Civil Liberties Union (ACLU), which opposes the profiling of air travelers by whatever means. “We don’t have enough information to create that profile.”
The U.S. Transportation Security Administration (TSA) denies that travelers are targeted for screening on racial or religious grounds. It has also abandoned attempts to create algorithms that would create non-racial profiles to identify particular travelers as posing a heightened risk. Instead, the TSA is trying to improve the existing system, which always targets for extra screening people whose names have been placed on a “selectee” watch list.
Journal reference: Proceedings of the National Academy of Sciences (DOI: 10.1073/pnas.0813202106)
