Holum properly stated the essence of the theoretical problem introduced by missile defenses: "For more than 30 years the classic argument in favor of strategic stability and against the deployment of a large-scale strategic missile defense system has been based on concerns that one side might have the ability to make a surprise disarming first strike against the enemy and then deploy a broad strategic missile defense system to knock out the enemy's combat resources which had survived the first strike and were being launched against the assailant."

In his subsequent claims, however, Holum seriously misstates the adverse effects of U.S. national missile defense on Russia's strategic deterrent:

Point 1: Holum says that the U.S. national missile defense (NMD) system is very limited and could protect only against a threat from a few dozen warheads; In contrast, Russia would possess between 1,000 and 2,000 warheads under START III over the next decade and thereafter. He concludes that such an abundant force will give Russia "the certain ability to carry out an annihilating counterattack on the other side regardless of the conditions under which the war began."

Comment: In reality, a surprise offensive U.S. strike could, under some conditions today, destroy all but a few tens of Russian warheads, and national control over those surviving weapons might be lost. [See the discussion below of estimated numbers of survivable Russian warheads.] In the future (2010-2015), the size of the Russian force could easily drop below 500 warheads, contrary to Holum's projection, in which case the protection afforded by a "very limited" U.S. NMD system would loom much larger in Russia's estimation.

Point 2: Holum refers to the Russian strategic arsenal's constant alert status and to its perpetuation under START III as grounds for the Russians' confidence that they could mount an annihilating counterattack.

Comment: Holum's observation suggests that Russia's alert posture is a good thing, when it actually represents a bad thing in terms of the accompanying risks of mistaken or unauthorized Russian launch. The U.S. position should seek to reduce, not embrace, Russia's readiness to launch on warning.

Point 3: Holum asserts that the Russian strategic forces that would survive a U.S. strike could deliver a minimum of a few hundred warheads, and that their sophisticated decoys would allow Russia to dispense with the traditional requirement of allocating Russian warheads to the mission of exhausting the defensive resources to overcome them.

Comment: This estimate of a few hundred surviving warheads is higher than Russian planners could estimate. Hamstrung by economic and other limitations on deploying its mobile sea- and land-based missiles into the sanctuaries of the oceans and forests, Russia is forced to leave its nuclear forces like sitting ducks in their garages, silos, and ports. Today Russia struggles to keep one or two ballistic missile submarines at sea, and one regiment of land-mobile missiles hidden in the field.

Comment: This estimate of a few hundred surviving warheads is higher than Russian planners could estimate. Hamstrung by economic and other limitations on deploying its mobile sea- and land-based missiles into the sanctuaries of the oceans and forests, Russia is forced to leave its nuclear forces like sitting ducks in their garages, silos, and ports. Today Russia struggles to keep one or two ballistic missile submarines at sea, and one regiment of land-mobile missiles hidden in the field.

All the rest of its strategic forces are vulnerable to quick destruction. The surviving weapons might consist of one submarine (48 warheads for a Delta III or 64 warheads for a Delta IV), and one regiment of SS-25 mobile land-based missiles (9 warheads). Depending on the effectiveness of U.S. anti-submarine operations against Russian boats (a routine activity still today), and depending on the extent of disruption of Russian command and communications, it is possible that a very small number of Russian warheads would be available to fire at targets in the United States, and that they could be neutralized by NMD.

Furthermore, Holum's estimate--a minimum of a few hundred deliverable Russian warheads--is not an acceptable number of surviving weapons from a Russian standpoint, just as several hundred surviving U.S. forces would not be acceptable to the United States. As a point of reference, the United States currently requires its strategic forces to be able to destroy in retaliation to Russian attack the vast majority of the nearly 3,000 targets assigned to them. (The number of targets in the U.S. strategic war plan actually grew by 20 percent over the past five years.) In other words, the United States must be able to deliver about 2,000 warheads in retaliation in order to perform the nuclear wartime mission to its satisfaction.

Holum also errs in presuming that Russian nuclear war gamers and planners would be less conservative than their U.S. counterparts in assessing the effectiveness of decoys.

Risk-averse planners on both sides credit the opposing missile shields with higher effectiveness than they deserve, and seek to overcome defenses by throwing more warheads at them than the defensive interceptors can possibly engage. The offense seeks to exhaust the defense's resources. For example, as late as the 1990s, 69 U.S. nuclear warheads were assigned to attack a single above-ground radar station in the ring of Russian interceptors around Moscow, even though the Russian interceptors would almost certainly have failed to destroyed many of them. Holum fails to recognize that if the Russians employed the traditional tactic of trying to exhaust the 200-interceptor defense envisioned in the plan he presented to the Russians, they would fall far short of doing so. Their surviving arsenal would be depleted trying to overcome U.S. defenses, and they would credit the U.S. defenses with the ability to negate their small retaliatory force.

Point 4: Holum tries to reassure the Russians that their ability to launch their alert forces as soon as early warning sensors detect a U.S. strategic missile strike in progress--launch on warning--would make it highly unlikely that a potential aggressor would ever contemplate initiating such a strike. He rightly notes that successful Russian launch on warning could dispatch about a thousand warheads, which would neutralize the effectiveness of the assault.

Comment: However, Holum surely knows full well that the Russians' confidence in their satellite- and ground-based sensors has declined dramatically over the past decade. Gaping holes in Russia's early warning network, which the United States would exploit in wartime, mean that the Russians must heavily discount the feasibility of this quick-launch option, even though they will strive to exercise it in a crisis.

What is more distressing is Holum's implicit calculation that the benefits of Russia's hair-trigger alert stance (which supposedly relieves its apprehension of U.S. missile defenses) outweigh the liabilities (inherent danger of mistaken or illicit firing, a risk that is growing as a result of deteriorating Russian nuclear early warning and command systems). What is even worse, Holum's positive view of this dangerous posture encourages Russia to maintain it indefinitely, when the most urgent priority today is just the opposite: to get Russia and the United States to take their strategic forces off high alert (de-alerting) in order to buy a larger margin of safety against a catastrophic failure of command and control.

Point 5: Holum further argues that in a period of increased international tension, Russia could disperse additional forces into the seas and oceans and thus augment considerably Russia's second-strike arsenal. This is partially accurate as far as it goes. Holum has not taken several factors into account, however. First, Russia's capacity for surging forces on short notice in a crisis has declined substantially over the past decade, for the same reasons that its combat readiness has declined in normal peacetime circumstances over this period. Second, the surging of strategic forces would be provocative and de-stabilizing in a crisis, and the temptation to do so would be greater today because of the vulnerability of Russian strategic forces in their normal peacetime configuration. Last, Russians view the resilience of nuclear command and control as far more decisive than the survivability of the individual weapons in the field, and their confidence in their command system in its peacetime or wartime configuration has declined sharply. (For instance, their deep underground command posts are crumbling and flooding.)

Point 6: Finally, Holum offers the following reassurance: "The tremendous risks associated with initiating a nuclear war under any circumstances make these theoretical calculations largely irrelevant. Obviously, neither side could ever contemplate such an assault."

They nonetheless prepare such an assault, and it hardly needs repeating that both Russia and the United States base their nuclear policy on each other's capabilities, not on their intentions. The inconvenient truth is that both sides are planning for the contingency of all-out nuclear attack.

Late in 1999, I visited a Minuteman launch crew in Wyoming and observed them going through the motions of launching their missiles at Russia. Nothing much had changed since my stint as a Minuteman launch officer in the 1970s. Two young men in their early to mid-twenties pulling alert duty just as I did, for the same expressed reasons, prepared to follow the same basic launch procedures that I learned nearly 30 years ago. Similarly, I spoke recently to the flight crew of a Strategic Command reconnaissance aircraft who recounted their recent foray along the northern coasts of Russia to find the holes in Russia's air defense radar through which U.S. strategic bombers could fly unscathed in the event of nuclear war.

With all due respect to Holum, the Russians are not consoled by his sincere belief that neither side would ever contemplate initiating a nuclear strike. The Russians pay more attention, to put it mildly, to the fact that if a launch decision were made right now in the White House and Pentagon, the order could be carried out within about two minutes. Those young men in their launch centers would validate the order, retarget their missiles (overriding the 1994 Clinton-Yeltsin de-targeting pact in seconds), and send the launch signal to about 2,000 strategic warheads, which would immediately fire out of their silos. About 10 minutes later, another 400 warheads would break water from the U.S. ballistic missile submarines on launch-ready alert at all times.

Bruce Blair is president of the Center for Defense Information