I have no idea why somebody as smart and knowledgeable as Nate Silver is using simple probability, a tool handy for determining the likelihood of the occurrence of random events, in analyzing the likelihood of acts of terrorism onboard commercial airlines. It’s an activity about as suited to the task as doodling cars on the back of a napkin is to winning the Indianapolis 500.
Terrorist attacks aren’t random and don’t behave according to the laws of simple probability. Mathematically speaking, they are games with players, strategies, costs, rewards, and multiple moves, each of which is dependent on previous moves.
If it is an attempt to say that the risk of such an event is low, that’s ignoring the moves that have already been made or assuming that the present strategy will continue to be effective forever regardless of the opponents’ moves. What do you think would happen if we eliminated all screening and security of all sorts on air travel? I think that by doing so we would have reduced the cost of an attempt and, as long as the rewards remained, we would see increased attempts and we would incur substantial costs.
What do you think would happen if we continued playing the same strategy while our opponents’ continued to alter theirs? That’s pretty panglossian in my view. Basically, it’s imposing a fixed cost on attempts and, ultimately, turning the game into one of imperfect information on our part and perfect information on the part of our opponents. The implications of such a situation are well know. If we failed to alter our defensive strategy in response to new developments, once our opponents had found a working strategy, we would see an increase in attempts and we would incur substantial costs.
Contrariwise, I believe that, once you’ve factored in the incentives, costs, and rewards, the only rational strategy open to us is to 1) impose some level of security; and 2) alter our security system as our opponents change their attacks.
This is not to say that I believe that we have made the right moves so far or that I think that the Travel Security Administration’s response to the most recent attempt was appropriate. As I have said repeatedly, I think our security system should have incentives and costs aligned and should play to our strengths rather than allowing our opponents to use our strength against us. It should be de-centralized, non-political, layered, and adaptive. That means that the federal bureaucracy is distinctly unsuited to the task of providing airline security.
So far we’ve done practically everything wrong but there’s still ample opportunity to do better.