« Yikes (or Foiling a Splash in the Pond) | Main | Zoom Zoom Zoom »

August 10, 2006

Searching for the Bad Guy

Thinking about the thousands of innocent lives that could have been lost in an attack on only nine 747s in the latest, foiled threat, some perspective might be in order. The image below shows all non-military flights being tracked in (or near) US airspace as of this evening at 9:25 MDT.

That there is a LOT of planes. A lot of people. A lot of lives.

I'm not saying that losing nine or ten planes would be inconsequential; certainly not: it would rival or surpass the horror of 9/11. I am saying that asking every new mother on every one of these flights to take a sip of their carry-on breastmilk or formula before allowing them on the plane is, well, silly. Non-productive. Approaching the problem from the wrong angle.

Something the TSA likes to do.

Luggage is not the problem. Carry-on items are not the problem. People are the problem. Certain people.

I'm not even going to go all wingnutter on you and suggest we target brown-skinned Arabs and give a pass to everyone else. Rather, it seems to me, that we need to focus on computer-assisted passenger screening (CAPS) and allow the systems to tell us for whom we should be looking:

In fact, in 1998 a U.S. aviation official, Cathryn Flunn, told a congressional hearing that computers can do this kind of figuring better than human beings....

Into this first computer-assisted passenger screening (or CAPS) system went some two dozen different piece of information about each passenger, taken from the airline computers. Out came a score, which, if it was above a certain number, caused the passenger's luggage to be "targeted" for a search....

No CAPS criteria, testified Flynn, "none of them, have to do with ethnicity of people; none have to do with association with names, with ethnic groups; none have to do with race or religion."¹

Does this work? Based on results from the U.S. Customs Service, the answer is "yes:"

Data soon confirmed that well-targeted methods for creating truly good statistics about real people could spot more criminals than do human-kind predictions based on race, ethnicity, or gender. In 1999... the Customs Service searched 23,108 people and found smuggled [contraband] in 533 instances. In the first eight months of 2001, the service had logged 734 seizures -- but it only searched 6,111 people.

So, a 50% increase in seizures with a 75% reduction in searches.

And minorities were no longer disproportionately targeted.

And there it is. CAIR won't be able to whine about the oppression of the Muslim brotherhood.

I'm a civil libertarian, and I'm naturally skeptical of the government acquiring personally-relevant information without a clear explanation of its intended use, and without explicit safeguards to prevent abuse. However, our energies are better spent discussing that topic than sniffing every bottle of shampoo going through Gate 5.

Although, that chamomile shampoo sure smells nice.

Update: The Wall Street Journal says just such a change in approach is taking place.

¹ Berreby, David. Us and Them: Understanding Your Tribal Mind. Little, Brown and Company, New York, 2005. pp. 53-55.