C.D. Harris (Ipse Dixit) rightly takes the administration to task for inserting an operational readiness testing exemption for the national missile defense system in the FY2004 budget proposal. C.D.'s more interesting comment is:

I have never understood - and I have tried - the Left's deep-seated loathing of missile defense. Why a purely defensive system - one which would exist for no purpose other than to protect American cities from nuclear annihilation - engenders such a consistent and emotional rejection from the peace, love and u-u-understanding set is, frankly, beyond me.

Me, I'm pretty agnostic on national missile defense. A 100%-effective NMD system (capable of stopping a large strategic missile attack) would be incredibly destabilizing to the nuclear balance of power, because it would effectively neutralize the threat of mutual assured destruction; at some level, this is probably the origin of the left's opposition (although it being a cornerstone of Ronald Reagan's defense policy probably didn't hurt). Of course, this isn't what the administration is planning; they want something that would stop a limited attack (say, no more than a dozen missiles), in essence to neutralize the first-strike threat by a minor nuclear power (say, North Korea).

Bret, from C.D.'s comments, makes two arguments against NMD:

1. It doesn't work. Scientists have said for years that the technology is barely even in its infancy, and won't be mature for at least 10 years.

2. It's incredibly expensive. I can't give you the exact numbers right now, but it's in the hundreds of billions.

To point 1: the technology won't even be developed if there isn't an impetus for it. So if we decide in 2010 or 2100 or 3000 that we want national missile defense, we still need to wait 10 years to develop it.

To point 2: it is incredibly expensive. Is it worth, say, $200 billion to ensure that Kim Jong Il can't nuke Honululu, Seattle, or Los Angeles? Or to ensure that a nuclear-armed fundamentalist Islamic state (say Pakistan or Iran) can't wipe out Diego Garcia or London?

Now, I don't know the answer to those questions. It's fundamentally a risk analysis.

On the other hand, there are far less expensive, but equally effective, delivery systems for nuclear weapons; for example, one could be loaded in the hold of a container ship or commercial jetliner, for example. National missile defense would be ineffective against the former, and there are other, cheaper methods (traditional SAMs, fighter jets) for dealing with the latter, if they are detected in time.

The bottom line: NMD alone would be an ineffective strategy. But, NMD might be a part of an effective overall strategy (also including improved human intelligence and signals intelligence) against WMD, if it can be developed cost-effectively, and there might be some civilian side-benefits to developing NMD (such as improved pattern-recognition systems, more efficient lasers, and better geolocation capabilities).