Commentary: It's Rocket Science--and That's Not Good

By Stan Crock

When a top State Dept. official recently told some Europeans about Bush Administration plans for missile defense, he made a revealing concession. He didn't claim that the new, scaled-back Stars Wars scheme would provide airtight protection against a handful of incoming missiles from a rogue state. Instead, the goal of a missile-defense system that would cost $120 billion, he said, was to complicate "a prospective opponent's calculation of success, adding to his uncertainty and weakening his confidence."

There's a certain logic here. Compared with Ronald Reagan's Star Wars, President Bush's limited antimissile scheme seems more defensible morally: It aims to protect people rather than missile silos. It also has one technological edge. In theory, at least, it's easier to hit a few incoming missiles than to destroy the thousands of warheads that Reagan's version was designed to stop.

NO COPILOT. However, the diplomat's admission highlights a critical flaw in the Bush team's game plan: While this downscaled missile defense may be feasible, it never will be reliable. Why not? Because tricky rocket and computer science are involved. Experts say an interceptor has a 70% chance, at best, of hitting an incoming missile. So the Bush team is counting on several opportunities to hit the missile--just after it's launched, in mid-course, and in its descent. But if each interceptor has a 70% chance of success, you never reach 100%. Supporters say that's better than no chance. But if the shield prompts China to expand its nuclear arsenal, and India and others follow suit, the net threat to the U.S. may grow.

In fact, launching missiles to knock out incoming warheads is a dicey proposition from the get-go. Aerospace leaders Boeing Co. and Lockheed Martin Corp. have lost plenty of rockets in lift-off, as have the Russians, the Europeans, and the Chinese. These projectiles aren't like ordinary planes, where a copilot takes over if the pilot passes out. In the language of complexity theory, rockets are "brittle"--one little thing goes wrong, and it's kaput. In last summer's test of the missile-defense system, a payload didn't separate from the booster. That glitch wasn't even on the Pentagon's list of potential worries.

And think about the software involved. How long does your PC stay up without crashing? The programs required for missile defense will consist of millions of lines of code. They'll be vastly more complex than a Microsoft spreadsheet or Netscape browser. There is no perfect method to debug programs of that size.

While we struggle with these challenges, the very threats we seek to thwart may prove ephemeral. The regimes in Iran, Iraq, and North Korea could fade away or turn benign. China, on the other hand, is bound to expand its arsenal. It certainly won't stand by and let its small retaliatory capability be blunted by our defense system.

A NEW MAGINOT? What's more, as we spend ourselves into oblivion on mini-Star Wars, potential adversaries will turn the missile shield into a high-tech Maginot line. The shield won't defend us against cruise missiles or suitcase bombs. And however strong a missile-based system is, offensive weapons can easily overcome it. As proof, consider North Korea vs. the U.S. In one corner, a tiny, poverty-stricken country manages to come up with an arsenal of attack missiles. In the other, the world's most powerful country spends $60 billion over 20 years and still lacks real defensive capability. Missile defense will raise the ante. But it won't alter any of the odds.

Crock covers defense from Washington.

    Before it's here, it's on the Bloomberg Terminal.