At 12:01 PM 10/28/02 -0800, Tim May wrote: ...

By the way, there are perfectly good fixes to the current hysteria about things carried on board planes. Besides the obvious absurdity of issuing alarms when fingernail clippers are found (but ignoring razor sharp edges in things like laptops with metal cases!), there are many fixes which can be applied:

I think the best fix is to accept that a determined suicidal attacker will probably manage to bring down the plane, but make sure that's the worst he can do. That removes the externality problem. The current algorithm for this is some combination of pilots being told not to go along with hijackers' demands, and maybe some chance of getting a military jet in place to shoot the hijacked plane down, if it is taken over by the hijackers. (It seems like this wouldn't be practical most of the time, e.g., if someone takes over the plane as it's approaching landing, there probably wouldn't be anyone in place to shoot in time. And faster response time means less time to discover a mistake.) I've heard of an idea for a mechanism for putting some kind of remote-control piloting mechanism on the plane, so that it can be taken over from the ground. This adds new attack points, but it might be workable. And of course, rockets have long had self-destruct mechanisms; presumably, there's stuff off the shelf from NASA or the DoD that does this with some reasonable level of security. (This last one would be politically unacceptable, but it's not really all that different from having a fighter shoot the hijacked plane down.) Both of these introduce a bunch of new vulnerabilities, though. Your list left out the obvious technique, which I think is more-or-less used by El Al: Screen your passengers really well, probably using secret databases, various kinds of racial profiling, etc. Routinely turn passengers away, or make boarding the plane such an ordeal that they elect not to fly anymore. (One of the many problems with this is that most flights are within the US; make flying sufficiently nasty, and people will take trains, busses, or their own cars. I think this is already happening a great deal, which is one reason most airlines are doing so poorly.) ...

4. Finally, market solutions are usually best. Any of the above could be implemented. If customers feel safer with a different baggage policy, they'll pick it. ...

I can't imagine this being done in practice, but I wish it were. The problem *is* an externality, but not the one you pointed out. Politicians in office right now will be blamed if there's another hijacking. So if I choose to fly Allahu Akbar Airlines for the short security checking lines, I get the benefit, but part of the cost lands on incumbent congressmen and the President. And those incumbents, unlike most people who get stuck with such costs, have the power to do something about it. (Something pretty similar happens with the FDA, right? If you get the new cancer drug a year earlier, you get all the benefit (maybe you get to go on living); the FDA gets the added risk of their being some horrible side effect. So they force a different trade-off on you than you'd prefer.)

--Tim May --John Kelsey, kelsey.j@ix.netcom.com // jkelsey@certicom.com