The North Koreans are attention-seeking again, flaunting the possible launch of some sort of rocket that possibly might be the putative Taepodong-2 ICBM. And, apparently, the US government is thinking of shooting it down with their ballistic missile defence system. Well, system. Experimental test rig would be more accurate.
This is incredibly stupid. Fortunately for the world, the story is in the Moonie Times of WMD-to-Syria craziness fame, so the bullshit adjustment factor is quite large. But if we briefly suspend disbelief, what would be wrong with trying to shoot the thing down? For a start, the North Koreans are not going to point it, if indeed it's an ICBM, anywhere near the United States. That would risk bringing the real missile defence into play - that is, the nuclear deterrent. Not even Kim Jong-il wants to do that.
So not only would it be not very pointful, it would be indefensible in international law (like they care) - it's not illegal to have rockets, and it's not illegal to test-fire them through space. And if it's not aimed at you, it's not self-defence. This may seem a minor point, but there is a more important one in there - ever since the high Cold War, there has been something almost amounting to a norm of customary law that satellites are not interfered with as they pass over sovereign territory - even the other side's recce satellites. (John Lewis Gaddis used this as a pirmary example for his thesis that the Cold War was really the Long Peace) Space is therefore akin to the high seas, which is excellent news for anyone who uses weather forecasts, telecommunications, or satnav, to name but three. The US derives a lot of economic value, and far more military value than anyone else, from the free use of space. It would be very foolish to encourage interference with other people's space activities, as the technology needed to build an ASAT missile is now much more available than it was in the 1960s.
That aside, what would shooting it down achieve? It would certainly slap down North Korea, and set them building as many decoy warheads as possible. Not a great payoff. Unfortunately, the poor performance of the BMD system so far suggests it won't hit the rocket. And the consequences of trying and failing would be really desperate.
The reasoning behind a small "son of star wars" system is as follows: even if we can't hope to deal with a strategic-level nuclear strike, that threat is covered by MAD. The danger comes from proliferation, from small numbers of rockets launched by (presumably) deranged rogue dictators who don't care about being nuked back. (Can you see the logical flaw yet?)
Even if it's not perfect, some BMD would be nice because it makes it much less certain that an enemy missile will get through, and hence the enemy will be less likely to risk it. (If you can't see a logical flaw yet, you're probably Ann Coulter.) This is all based on the assumption that the other side is pursuing a minimal deterrence strategy. Minimal deterrence, as theorised by the French and Israeli defence establishments and by the UK with regard to the "supreme national interest" non-NATO role of its deterrent, argues that the degree of devastation brought about by even a small nuclear strike is so horrific that it is enough to deter anything less than a suicidal attacker - no war of choice in the Pacific would be worth Los Angeles or Seattle, to put it concretely. So, just a small force would be enough to provide credible deterrence for most purposes.
Clearly, this obviates the argument that the threats of today are irrational and therefore impervious to deterrence. BMD is meant to defeat minimal deterrence by making it uncertain whether or not the bomb would get through. Given that this time we know there is no bomb, it's very stupid indeed to run the risk of missing and thus destroying the uncertainty created by the BMD project.
Update: as always, any post with rockets in means comments action. Here's a nice little map courtesy of Chris "Chris" Lightfoot.
Update the second: here's an even cooler map courtesy of Armscontrolwonk.com