Juliet Lapidos writes a good Explainer column over at Slate about the fact that the TSA has never claimed to have stopped a terrorist at an airport checkpoint. But this is the most salient point:
What these numbers don't get at is whether the TSA airport screeners prevent terrorist attacks through their very existence -- deterring plots by hanging around. This is quite probably the case, but it's not obvious that they prevent any more attacks than the private contractors who handled checkpoints before the Aviation and Transportation Security Act of 2001 went into effect.
I think that the deterrence argument is pretty persuasive, but it has limits since a really determined terrorist can probably get by even the new security measures. The body cavity scenario for example, is far from unheard of. Which brings me to Nick Baumann's piece on this arguing that we have no idea whether this stuff is actually working:
It's not about doing something "instead" of the current system -- it's about not doing things that are wasting money and time and not making us safer. It's quite possible that we're already as safe as we're going to get -- and every subsequent airport security "improvement" is just reducing our freedom without improving security.
The problem isn't just that we don't know that; it's that the TSA doesn't either.