TSA Discriminates, But Inefficiently - Risking Lives and Wasting Resources, Says Law Prof

Dec. 30, 2009 - PRLog -- It is also illogical and nonsensical from a mathematical point of view, and unfair to many Muslim and Arabic travelers, says Prof. John Banzhaf, developer of the statistical tool known as the "Banzhaf Index," and recognized for his work in mathematics and game theory.

The operating procedures manual accidentally posted on the Internet by the TSA shows that the agency is now using both ethnicity and religion, at least indirectly, as factors in screening. The manual instructs screeners to automatically select, for secondary screening, all citizens of 12 named countries. Aside from 2 Communist nations, all of the others have large Muslim populations, including 8 which are at least 90% Muslim, and most are Arabic. This hardly seems to be just a coincidence.

But since all travelers from these countries are automatically singled out for secondary screening, the system fails to distinguish between toddlers or elderly women from the suspect country on the one hand, and young males from these same countries on the other.  Thus elderly women, young children, and others from these countries very unlikely to pose any threat are automatically singled out simply because of their country of origin; virtually all Muslim and most Arabic.

This system also would not have singled out Umar Farouk Abdulmutalib for secondary screening, even though he is a young male from the 4th most populous Muslim country, suspiciously used cash to buy a one-way ticket, was on a watch list, and chose a seat known to make the aircraft especially vulnerable to explosions.

This is important because explosive materials, and bombs without metallic detonators, do not set off the metal detectors all passengers are required to walk though, and the more thorough secondary screening techniques which may be able to detect them -- e.g., using whole-body imaging scanners, "puffer" bomb-detection machines, or a lengthy hand pat down -- are either too expensive and/or take too much time to be used on the majority of passengers.

Thus logic, and the mathematics of testing and inspections, dictates that the only efficient method is to largely single out for secondary inspection those passengers most likely to be terrorists, and not to select fliers for the more thorough secondary inspection completely at random as the TSA now claims it does.

One way to help identify passengers who are substantially more likely to be terrorists is to utilize, in combination with other recognized signs, factors like nationality, gender, and age -- and possibly even religion and ethnicity -- in selecting passengers for secondary screening.  Both the Supreme Court and the Justice Department have indicated this is constitutional, provided only that no single factor is decisive -- as it unfortunately is under the current TSA system.

Increasing the probability that those singled out for secondary screening are terrorists by considering ethnicity and/or religion, as well as gender, age, how, when and where the ticket was purchased, the demeanor and clothing of the passenger, etc., is the only logical way to make secondary screening effective and efficient, and capable of stopping terrorists without unreasonably delaying the traveling public or requiring enormous expense for new equipment and TSA screeners.

While some have suggested that behavioral analysis -- a technique pioneered Israel -- is preferable, it seems unlikely that this would work efficiently in the U.S. with our current TSA screeners and huge number of passengers.  Moreover, news reports suggest that Israel does in fact use some profiling techniques, logically giving more attention and scrutiny to young Arab males than to elderly Asian females or Jewish-American children.

"We readily accept the proposition that persons buying one-way tickets for cash, or those apparently wearing more clothing than necessary, should be subject to greater scrutiny, even though only a tiny minority may be terrorists, and the predictive effectiveness of these factors may be somewhat limited.  It may be time to do the same for religion, ethnicity, gender and age, which may be at least as effective as predictors.  Let's stop giving young Muslim males no greater scrutiny than elderly Asian women, just because it seems politically correct," says Banzhaf.

PROFESSOR JOHN F. BANZHAF III
Professor of Public Interest Law
George Washington University Law School
FAMRI Dr. William Cahan Distinguished Professor
FELLOW, World Technology Network
2013 H Street, NW, Washington, DC 20006, USA
(202) 659-4312 // (703) 527-8418
http://banzhaf.net // http://ash.org