Superstition at the Checkpoints

Last week, Rob Nordland filed a great story about the Iraqi police's use of the ADE 651, a bomb-detecting device that costs "$16,500 to $60,000 each" (love that margin of error) and does not, strictly speaking, detect bombs. The people at the James Randi Educational Foundation, never ones to decline a bet on a sure thing, offered a million dollars to the manufacturer if it could prove the device worked better than chance. The manufacturer, based in London, has not taken up the challenge, and the overall impression of the sneering article is that the Iraqi security forces are being blown up, and allowing their countrymen to be blown up, because they are too scientifically illiterate to know they've been had.

I'm sympathetic to Andrew Exum, who calls the deployment of these stupid toys "deadly idiocy." But Mauro De Lorenzo writes in to flag this passage from Nordland's report, and suggest a parallel:

Proponents of the wand often argue that errors stem from the human operator, who they say must be rested, with a steady pulse and body temperature, before using the device. Then the operator must walk in place a few moments to "charge" the device, since it has no battery or other power source, and walk with the wand at right angles to the body [...]. If, as often happens, no explosives or weapons are found, the police may blame a false positive on other things found in the car, like perfume, air fresheners or gold fillings in the driver's teeth.

Scientifically, this is preposterous. But who can deny that the prerequisites the Iraqis name for the proper operation of the device will make for better bomb searches? This magic wand (it is nothing less) forces Iraqi police to keep cool, to meditate themselves into a confident trance, and to approach suspicious vehicles with authority. Call it the Stone Soup theory of checkpoint bomb detection.

De Lorenzo offers an analogy to a Mayi-Mayi fetish, which stops bullets, but only if you follow a number of rules that themselves improve the chances that the bullet will miss you, perhaps by giving you enough confidence to evade the shooter, or even spooking him into missing on purpose. I would certainly shoot a little to the side, if I thought my enemy had magic on his side. Likewise, from the perspective of the potential bomber (bombers who do not subscribe to the James Randi blog, anyway), imagine having an Iraqi policeman approach, confident and armed with expensive equipment -- will you be more nervous, or less? Will you sweat more, and grip the steering wheel a little tighter? If so, the ADE 651 may have worked despite itself.

Finally, consider an American parallel. This magazine has long since concluded that the TSA searches are farcical and intrusive. When we hear that screeners miss bombs and guns in 20 of 22 tests, it is not much better than chance, depending on how you count. But does this vast and annoying security apparatus deter possible bombers? I suspect it does, no matter how Clouseau-esque it appears to careful, rational observers. It has a comically high false-positive rate, but few of us ever have the experience of bringing bombs onto planes without detection, so the false-negative rate is not something most of us can casually observe. (I exclude the false-negative rate for jackknives, Dasani bottles, and tweezers -- objects that no terrorist would ever bother with anyway.)

More technically speaking, to lay observers the screening at airports has low specificity but undetermined sensitivity. Less technically speaking, airport screening is a load of hooey. But we do it anyway, and one reason may be that having the procedures calms the nerves of those who are watching for telltale suspicious behavior, and inflames the nerves of those who have reason to be nervous. I have no idea if this hypothesis holds true in the Iraqi or American case. But if it does, I can understand why Iraqis and Americans would be reluctant to come clean and admit that their security works largely because of the false illusions on the part of terrorists and screeners alike.