As the history of the cold war clearly demonstrates, nuclear primacy is a notoriously poorly defined and elusive goal. This is why it was very interesting to read the article by Keir Lieber and Daryl Press, "The Rise of U.S. Nuclear Primacy", in the March/April issue of Foreign Affairs. The authors argue that the United States has achieved (or is just about to achieve) nuclear primacy, which would make the concept of mutual assured destruction obsolete by taking "mutual" out of it. The United States, the authors argue, is getting the capability of destroying all (and they mean literally every single one) Russian strategic launchers in a first counterforce attack. A longer article, which the authors promise will contain details of the model they used to calculate effectiveness of a first strike, will appear in the Spring issue of International Security.

The argument looks convincing at first - Lieber and Press paint the familiar picture that many now take for granted - decline of the Russian strategic forces, gaps in the early-warning network, and all that. But a closer look at the article reveals a number of factual errors and unsubstantiated statements, which in my view completely undermine its conclusions.

Let's start with how Lieber and Press make their case when they try to prove that the Russian strategic arsenal "has sharply deteriorated". They quote the START Treaty statistics that tells us that "Russia has 39 percent fewer long-range bombers, 58 percent fewer ICBMs, and 80 percent fewer SSBNs than the Soviet Union fielded during its last days." This sounds about right (with the exception of the number for bombers - in 1990 the Soviet Union had 162 strategic bombers and Russia has 78 of them now, making it a somewhat steeper decline of 52 percent). But these numbers hardly prove anything. If we look at the U.S. forces, we will see that during the same period the number of missiles in the U.S. ICBM force went down from 1000 in 1990 to 500 today - a 50 percent decline. The United States "lost" two thirds of its strategic bomber force - it is down from 347 in 1990 to 115 in 2006. The U.S. Navy, as one would expect, preserved most of its strength, but even there the number of submarines is down to 14 from the 1990 level of 33, a decline of 58 percent. As we can see, the numbers on the U.S. side are pretty close to the Russian ones. The reason is simple - these numbers reflect a completely different process - reductions of nuclear arsenals from the insanely high cold war levels. To use them as evidence of force deterioration is quite misleading.

Of course, this does not automatically mean that the Russian forces are in good shape or that their evolution during the last fifteen years can match the modernization program undertaken by the United States. But this evolution is a much more complex process than the one presented by Lieber and Press.

Strategic aviation, traditionally the weakest leg of the Soviet and Russian triad, has received quite a boost lately - only in 2005 it participated in four major exercises. I guess some would still take this as a proof that strategic bombers "rarely conduct training exercises", but I'm not sure this statement accurately describes the state of affairs today. The strategic aviation has quite ambitious plans for this year, in terms of exercises and bomber modernization program. Yes, bombers are vulnerable to a surprise attack, but it is unclear how this vulnerability is greater today when they are "located at only two bases" and not at four, as it was the case before the Soviet breakup. Equally unclear is since when bomber warheads "are stored off-base", as Lieber and Press claim. To tell the truth, none of this is really relevant to the discussion of first-strike vulnerabilities, since neither the Soviet nor Russian military would ever assume that bombers have any chance of surviving an attack. But if the purpose of giving us all these details about the Russian strategic aviation was to demonstrate that it is in bad shape, then the "evidence" does not look entirely convincing.

The same is true for the description of the Russian strategic submarine fleet. While it is indeed in a difficult situation, it would be wrong to focus on comparing patrol rates of U.S. and Russian submarines. These navies operate in completely different modes and have entirely different missions. The sharp decline in patrol rates of Russian strategic submarines does indicate some loss of capability, but it is not nearly as serious as it would have been the case of the U.S. Navy had it suffered a similar decline.

To further make the point about the decline of the Russian navy, Lieber and Press refer to a number of failures during recent missile tests. But their evidence is misleading at best. They mention the two launch failures in February 2004, but forget to point out that the missiles involved in those tests were older modifications of the R-29RM/SS-N-23 missile. These are now being replaced by a newer modification of this missiles, known as Sineva, which just successfully completed flight tests and was equipped with a new warhead. The launch failure in summer 2005 involved a Volna launcher, which is an old R-29R/SS-N-18 missile modified to carry payload into space. In the other alleged failure, in a launch in October 2005 of a "Demonstrator" reentry capsule, it was the capsule that failed, not the missile (although that launch was postponed at least once because of the problems with the missile). While these incidents do tell a story about reliability of old submarines and missiles, it is hardly something we do not already know - that old submarines and missiles are getting older. This is nothing new and these submarines are slated for dismantlement and unlikely to remain in service much longer anyway.

What Lieber and Press do not mention when they describe SLBM launches is a number of successful tests of R-29R and R-29RM missiles and the apparently successful Bulava missile development program. Two successful flight tests of this new missile conducted in September 2005 and December 2005 indicate that the Russian strategic fleet will eventually receive the long-awaited Borey submarine, even though it is unlikely to happen before 2008-2009. Until then, Project 667BDRM/Delta IV, which are undergoing overhaul, and even some Project 667BRD/Delta III submarines would keep the Russian navy in business.

Again, this is not say that the Russian strategic fleet will ever be able to match the U.S. Navy. But by its own, much more modest, standards the Russian fleet has been doing quite well recently and seems to be quite adequate for the task that the Russian doctrine has for it.

An assessment of the situation with land-based missiles also requires close attention to details. First of all, one should be careful with the concept of "service life". While it may be true that "over 80 percent of Russia's silo-based ICBMs have exceeded their original service lives", this number hardly tells anything. For most missiles the "original service life" was set at the level of seven to ten years, but then was extended as the Rocket Forces accumulated experience in operating the missiles and the designers grew confident in their performance. Regular successful tests of missiles that were kept in silos for 25 years or longer confirm that this extension does not necessarily reduce reliability of a missile. Today, the Rocket Forces confidently discuss the possibility of keeping missiles in service for 25-30 years or even longer. Most of the missiles deployed today can easily be kept in service for five to ten more years and if Russia is decommissioning some of them it is because it does not need them, not because they are unreliable.

To replace the old missiles, Russia has been developing an deploying a new missile, Topol-M. Development of this missile has a long and complex history, but it is highly inaccurate to say that it was "stymied by failed tests". Of the fifteen flight tests of the missile conducted so far, only one failed - quite a remarkable achievement for any missile development program. As for the production rate, it may have been rather low - only 42 missiles have been deployed since 1997 - but the question to ask here is, low compared to what? A good case can be made that Russia does not really need silo-based Topol-M and should rather concentrate on building the mobile version of the missile. Which is exactly what is happening - the first regiment of mobile Topol-M missiles is expected to enter service later this year.

Yes, it is true that by the end of the decade and later Russia may have as few as 150 land-based missiles, which is a big step down from the 1398 it had in 1990. But Russia does not need 1400 missiles today and most likely will never need that many again. As for the 150 missiles, about half of them would probably be road-mobile Topol and Topol-M, which, if operated properly, would have a good chance of surviving a first strike. Lieber and Press dismiss mobile missiles by saying that they "rarely patrol". It is not clear, however, on what evidence do they base this claim. If there is some source of this information, they did not tell us what it is. In reality, very little is known about mobile missile patrol rates. There is some evidence that suggests that during normal peacetime operations no more than 25 percent of missiles are supposed to be in the field. It is quite plausible that patrol rates are somewhat lower these days, but it is a stretch to assume that they are zero.

It can be argued that land-based mobile missiles may be vulnerable even when they are on patrol, for the United States is constantly improving its capability to detect them. But even if this is true for missiles on peacetime patrol, locating the missiles would be much harder, if at all possible, if they get a chance to disperse in a crisis, as we discuss later.

Now let's turn to the Russian early-warning system, which Lieber and Press describe as "a mess". The problem with that characterization is that while the system is indeed past its prime, it has lost surprisingly little of its capability to do its job. This may seem counterintuitive at first, but Russia would gain very little were its early-warning system be deployed to the full extent. For example, adding the capability to detect launches of sea-based missiles, in North Atlantic or elsewhere, would not dramatically increase the time available to Russian leadership for attack assessment. The much talked about "gaping hole" in the radar network coverage that seems to exist in the east, also should be put into a proper context. First of all, the "hole" is not as large as it may appear - only missiles flying along carefully selected trajectories to a limited set of targets would be able to avoid detection. But more importantly, there is no way missiles launched from the Pacific could cover the entire range of targets to be destroyed in a first strike. A scenario in which "Russian leaders probably would not know of the attack until the warheads detonated" is a theoretical construct that has nothing to do with reality.

Speaking of theoretical constructs, the counterforce first strike that Lieber and Press describe assumes a bolt out of the blue attack, done with no preparation and without provocation. But hardly anyone ever believed that an attack of this kind is possible. Certainly not anyone in the Soviet Union or in Russia, which assumed that any attack would be preceded by a period of tensions, giving them the time and opportunity to disperse the strategic forces and bring the command and control system into a high degree of readiness. This is quite a reasonable assumption I should say. These measures would be relatively easy to implement and while they may not make the forces completely invulnerable, they would significantly raise the uncertainty associated with the first strike.

There are other serious issues with the paper. I believe that Lieber and Press seriously overestimate the potential utility, whether military or political, of counterforce first-strike capability and misread the history of the cold war. For example, never during that time the Soviet Union had "the hope of gaining nuclear superiority", even though it did end up building a massive nuclear arsenal. But getting into these issues would require a much longer reply and a different set of arguments.

The point of this message is that we should be careful with facts and their interpretation if we are making claims as bold as the one about rise of nuclear primacy and the end of MAD. Whether we like it or not, getting "mutual" out of MAD is much harder than putting it in.