Tomorrow, January 25th, 2005 is the tenth anniversary of the Black Brant XII rocket launch from the Andøya rocket range in Northern Norway. It would have been just another scientific experiment, but the rocket was quite different from all previous 600+ rockets launched from Andøya, so when the Russian early warning radars detected it, they did not recognize it at first and generated an alarm, which went up the chain of command. How far up did the alarm eventually travel is a matter of dispute, but it is believed that it made its way to president Yeltsin, who used his "nuclear briefcase" to communicate with the minister of defense.

With time, the story generated is own mythology and become an illustration of dangers of launch-on-warning posture and of the decline of the Russian early-warning system. And although there is evidence that the danger of that particular incident has been seriously exaggerated, there is no doubt that launch-on-warning postures of U.S. and Russian nuclear forces create risk of an accidental or inadvertent launch. Even if we can argue that the probability of thins kind of launch is extremely small, its consequences would be devastating and hardly anyone is ready to argue that this is the risk worth taking.

One way to deal with the dangers of accidental launch is to remove missiles from hair-trigger alert, associated with launch-on-warning postures. This approach, known as de-alerting, has been advocated for quite a while, but has not created a great deal of enthusiasm either in the U.S. or Russian government.

In my view, the reason for this is that the model suggested by advocates of de-alerting is unworkable and in some important respects would make the situation worse, not better. De-alerting itself is a very good idea, but in order to make it work, we need to reconsider a few things. I develop a more detailed argument in a memo submitted to a PONARS meeting in Washington in February, but the main conclusions are here:

First of all, we should realize that the U.S. launch-on-warning posture is potentially more dangerous than that of Russia and therefore removing U.S. forces from hair-trigger alert should be a high priority mission regardless of whether or not Russia is ready to reciprocate.

Second, efforts to repair the deteriorating Russian early-warning system should be abandoned as counterproductive. Attempts to repair the system or augment it would, in fact, increase the probability of accidental launch. In short, no system - no problem.

And, the last but not the least, attempts to make de-alerting transparent and verifiable should be avoided at all cost. De-alerting should not be treated as disarmament, where transparency and verification naturally belong.