The idea goes that some bad guys tried to blow up a TSA flight using liquids smuggled into a plane. This isn't a new idea, and has been successfully done before. Given that upgrading chemical sensors is expensive and time consuming, they implemented a temporary ban on all liquids. The temporary ban is now permanent. While this does appeal to the common senses, it is a bad security choice, as the overwhelming majority of liquids are not explosives, and there have to be some exception (see the TSA's exceptions list). Without advanced chemical sensors, all this ends up doing is increase the overhead of checking in with very little benefit.

It is essentially a poorly configured intrusion detection system (IDS) with more false positives that it can bear.

Schneier called this CYA security, because TSA just wanted to make sure that if a bomb went off again, it wouldn't be using an already attempted technique, something they wouldn't be able to evade responsibility for. This is silly because there are more possibilities for bomb creation, than there have been attempted, and checking for 'all' possibilities would be impossible. So TSAs, hidden corollary is that if a bomb goes off using a new technique, well then; "How can we predict a terrorist's innovation?"

ACSA has now added another layer[1] to the CYA model. I am sure they are implementing the ban, not because it is a real security benefit that provides measurable security improvements, but because TSA did it. You see, if a flight from SA gets blowed up with a liquid bomb, then people will shout at ACSA saying "Why didn't you do what TSA did?", and ACSA needs to cover their ass.

[1] Layered silliness, it's like defense in depth: Silliness in depth.