Two points that Mike emphasizes in his book struck me as especially surprising. First, he emphasizes again and again that we need to look not only at “worst case scenarios” and “ten-foot-tall terrorists” but at a full spectrum of more realistic scenarios. Second, as he puts it in his first post: “Here is how the process usually goes: someone proposes a defense; someone else points out a way that a terrorist group can get around it; everyone goes back to square one. (I exaggerate a bit, but it's close.)” I see the opposite of each of these phenomena as much bigger problems for the world’s response to the risk of nuclear terrorism.

Terrorists – ten-foot-tall or two-foot-tall?

In judgments about how much effort to put into securing nuclear stockpiles around the world, I repeatedly see gross underestimation, not overestimation, of likely terrorist capabilities. Some nuclear security “upgrades” I’m aware of were designed only to be able to defend against a “one plus one” threat – that is, one armed outsider, possibly with help from one passive insider. To protect nuclear bomb materials, I think it’s reasonable to say that’s not enough, even in the safest countries in the world. The manager of one HEU-fueled research reactor was extolling his security system to me (it included, among other things, a total of one police officer with a sidearm), and made the point that the power system was rigged so that if there was an intrusion the power would go out and the intruders would be in the dark and unable to steal the HEU. I inquired whether he didn’t think the adversaries would bring a flashlight. The Nuclear Regulatory Commission continues to take the view that terrorists won’t have rocket-propelled grenades – available in bazaars across the world. Now, a reasonable argument can be made that the post-9/11 security requirements at U.S. Department of Energy sites (which these sites are still struggling to come into compliance with), cost more to meet than is sensible to spend when other sources of nuclear material are so much more vulnerable – but over most of the world, the problem is defense against too little adversary capability, not defense against too much.

Similarly, a distressing number of people in charge of important aspects of nuclear security in different countries believe it is completely impossible for terrorists to make a nuclear bomb, even if with the right amount and quality of HEU. Pakistani President Musharraff’s famous remark to George Tenet that this is totally beyond the capability of “men in caves” is only one of countless examples. These beliefs fundamentally undermine the willingness of political leaders and nuclear managers to devote the resources needed to keep nuclear stockpiles effectively locked down for the long haul. (In Securing the Bomb 2007 I outline a number of steps that might be taken to make the case around the world that nuclear terrorism is a real threat and current nuclear security measures are often not sufficient; see pp. 103-107).

Imperfect defenses: rejection or over-investment?

Here, I would simply challenge Mike to provide a real case of the dynamic he describes – that is, where a proposed defense that he thinks was worth the money was rejected because it was imperfect. I see the bigger problem as throwing large sums – and a large amount of the political capital with other countries focused on this problem – on layers of defense that probably will not buy us a great deal. Congress’ rush to require that 100% of the cargo containers coming to the United States be scanned for radiation by 2012 – with absolutely no statement as to how good that scanning should be, no similar requirement for better detection for any routes other than cargo containers, and no system-level assessment of how useful this would be – is an obvious case in point. The obvious question about such systems is why terrorists would choose to route their nuclear material through crossing points with large and easily observable detectors in the first place. (William Langeweische, in his otherwise awful book The Atomic Bazaar, has a highly entertaining description of a multi-million-dollar crossing point in Georgia, built with U.S. assistance, that smugglers easily and routinely go around. You can check out my review of Langeweische’s book here.)

In his opening chapter, Mike offers a useful analogy of a baseball manager trying to figure out how much he should spend on which positions, given an overall salary cap. If we lived in a world where even the best outfield would only have a 5% chance of catching the ball, the manager might well conclude that he should effectively ignore players’ outfield capabilities and simply hire the best possible batters and pitchers. The fact that each layer of an integrated system has some chance of working does not tell us whether it is worth much effort and investment in comparison to other layers.

While Mike’s book provides a very useful system-level overview, I am concernced that the way the argument is presented ends up tilting the reader too much toward the lines of defense that are likely to offer less risk reduction per unit of effort invested, and too little toward securing stockpiles at their sources. Mike and I fully agree that keeping potential nuclear bomb material from being stolen in the first place “should be the foundation of any sensible approach to preventing nuclear terrorism,” as he puts it. But perhaps because he felt the need to make the case that this was not the only thing we should be doing, the entire chapter about security at the source focuses on ways that layer of defense might possibly fail, while the chapters on all other layers of defense focus on ways those layers might possibly succeed. Both of these kinds of discussions are correct, but this approach leaves an unbalanced impression of the weaknesses of security at the source and the strengths of later lines of defense. This is compounded by the fact that in an entire book on preventing nuclear terrorism, a total of one paragraph is devoted to what we should do next on securing nuclear stockpiles at their source (in case you missed it, it’s on p. 142). Mike’s feeling may have been that this didn’t need much more elaboration because it was obvious and others have elaborated on it at length (as I do most recently in Securing the Bomb 2007). But there are so many serious and difficult obstacles to be overcome in securing nuclear stockpiles, and so many steps that can and should be taken but have not yet been taken, that I found it surprising not to spend at least a little time walking through them – it creates, unfortunately, the impression that this part of the defense is reasonably well in hand, which is not yet the case. There is a lot more to be done – and urgently – to reduce the danger that nuclear weapons and the materials needed to make them could be stolen and fall into terrorist hands.

So, kudos to Mike for a major contribution to the debate – but in several areas, I would argue for a different balance.