Peer-review security

Typical plane: 200 passengers. Most of them want the plane not to crash, but occasionally one or two do.

Typical security system: Your ticket price pays for TSA employees to scan your bags. Each bag gets one pair of eyes (x-ray-machine-aided) on it for about 15 seconds. If there's a bomb in the bag, this is the only chance to catch it. The TSA fellow is well-meaning and well-trained, but presumably quite bored.

Peer-review security system: the passengers scan each other's bags, on the principle "many eyes make short work". Each bag gets searched twice, by two unrelated passengers. The passenger assignments are randomized by computer. If any search turns up something suspicious, the searcher pushes a button, and the bag gets a few minutes' going-over by a well-paid TSA professional.

1) 200 passengers, max 3 bags per passenger. Each passenger does six inspections. At two minutes each, everyone needs to arrive 12 minutes early---considering that the normal security clearance can take an hour, that's fine. I'm picturing open-the-bag-and-look inspections, but you could also hand the passengers x-ray scans for inspection. (This might not work without training.)

2) Let's take a plane with N passengers, of whom T are terrorists, with 1 bomb between them. The odds that the bomb-bag gets inspected once by co-conspirator are (T-1)/N. The odds of being inspected ONLY by co-conspirators is (T-1)(T-2)/(N(N-1)). Even if the passengers are poor inspectors, there's still a quite low chance of a bomb going undiscovered. You can raise the number of inspections per bag, depending on how effective a zero-training search really is.

3) You can imagine this being implemented by one airline/terminal at a time. Passengers on this airline voluntarily waive privacy on their bags, and get to skip a lot of the normal terminal-access rigamarole. There _does_ have to be some sort of personal scanning (metal detector, sniffer machine), because peer-review patdowns are not a good idea.