> Does that sound stupid? I hope so, because it is. But that's exactly the type of argument that this article is making. Actually, it's not just making that ar

> Does that sound stupid? I hope so, because it is. But that's exactly the type of argument that this article is making. Actually, it's not just making that argument, it's referencing studies to back it up: http://www.campbellcollaboration.org/lib/project/11/ http://www.campbellcollaboration.org/lib/project/11/ > At the base of this argument is an implication that terrorist attacks are zero-sum: reduce them in one place, get an equivalent more in another. But that's not how security works. The analogy you give certainly does sound stupid, but nobody said that the effect was necessarily zero sum, what the sum is is in fact an empirical matter and the research was about precisely this question. Unlike passwords, it would not be all that surprising for it to be approximately zero sum in this case. Terrorists don't sit down and think 'how can I attack an airplane', they sit and think 'how can I cause maximum damage for minimum investment'. Against that kind of opponent, there's no point choosing one tactic and fortifying it beyond reason while leaving a million other options for someone pursuing a strategy of terror. The effective money is spent on identifying the criminals and countering them rather than cherry picking one or two specific techniques. You can see this in most historical counter terrorism efforts. Very occasionally, it's effective to identify point fixes for things that are low hanging fruit for terrorists. Reinforced cockpit doors almost certainly were part of that. Almost everything else that's been done on air travel since? Not so much. > Second, it's pulling a subtle slight of hand--the goal of airport security is NOT to reduce terrorist casualties overall. It is to reduce violent attacks of any kind on airplanes, specifically. That's the job of the personnel at the airports. However, we want to know whether the TSA is a program that should be funded, funded at the level it currently is, or given the powers it currently is. That is a question about the most effective way to reduce terrorist casualties overall (or arguably limited to US citizens anyway). It's no sleight of hand to be concerned about efficiency in the measures we take as a society to combat terrorism. > First, you cannot draw such a causal conclusion from statistical data alone. It's not clear to me why you think that the burden of proof lies in that direction. We have a problem - terrorism. We take some expensive measures - over the top checks at airports. It's not "data wonky" to demand that the programs we've put in place are able to produce real evidence of effectiveness.