ABSTRACT
Ambassador Robert Gallucci is Distinguished Professor in the Practice of Diplomacy at the Edmund A. Walsh School of Foreign Service, Georgetown University, where he earlier served as the Dean. He led the John D. and Catherine T. MacArthur Foundation as president from 2009 to 2014. His career in the US Department of State spanned 21 years and included serving as Special Envoy to deal with the threat posed by the proliferation of ballistic missiles and weapons of mass destruction, Deputy Executive Chairman of the UN Special Commission (UNSCOM) overseeing the disarmament of Iraq and Assistant Secretary of State for Political-Military Affairs. He led the US negotiation of the Agreed Framework with North Korea in 1994. In this interview, he shares his critical views on the connections between nuclear deterrence, arms control, and missile defenses and suggests ways in which new technical capabilities might alter deterrence calculations. The interview took place on 5 November 2021 and was edited for clarity for the Journal for Peace and Nuclear Disarmament (J-PAND).
Discussions of deterrence and arms control sometimes portray them as opposites. Deterrence becomes synonymous with more numbers and better technology while arms control is about lowering those numbers. It’s a little more complicated, right? Some technologies are not always stabilizing or help arms control. How do you see the relationship of deterrence to arms control?
Deterrence is way of providing for security in a world in which we cannot deploy a defense by denial. We spent decades, even hundreds of years, building defenses as a method of providing security and then in one fell swoop at the end of the Second World War, we found ourselves in a world in which we had introduced nuclear weapons and the Germans had introduced the delivery vehicle, the V2 ballistic missile. And you put those two together and you have a weapon that cannot be defended against. Way back in 1945 we had no way of shooting down a V2 missile.
And in 2021 we have no way of shooting down, with any confidence, a sophisticated ballistic missile. And those nuclear weapons have only gotten bigger over the years, from kilotons to megatons.
I think early on we recognized that deterrence was going to provide our security and began to intellectualize over the concept of deterrence. It was “satisfactory” if it provided stability. That meant two kinds of stability at least: arms race stability and crisis stability.
If the relationship between countries is one in which they don’t continually have to race with each other to build more offense, then it’s stable; if they constantly have incentives to build more offense, it’s not stable.
The other kind of stability is more familiar to most people – crisis stability. This means that in the midst of a crisis, nobody has an incentive to strike first. Both sides have secure, survivable, retaliatory second-strike forces. So, if you have crisis stability, there’s no advantage to either side of a surprise attack. That means nobody strikes first. Now, this is a little bit wonky I recognize, but it is the best way to link our concept of security through deterrence with our concept of arms control as an essential management tool, so that we produce both kinds of stability – crisis stability and arms race stability.
And without that, to just simply trust to the good nature of man would be insane.
Man in a state of nature – which is the international system of states as we all know from Hobbes and a lot of other folks who have written about this – is not a nice guy. So we need to manage the instinct to defend ourselves in the state of nature, so that we don’t destroy ourselves.
Deterrence Is Our Only Hope but Essentially Unprovable
At the end of the Cold War, I think the United States assumed that it had arms race stability because the Soviet Union had collapsed and Russia and the US were engaged in a process of arms control. Most of the discussions I have been involved in have been more focused on crisis stability. Especially in third countries like Syria, where we had agreements to deconflict to avoid escalation. Where are we now? It looks to me like we don’t have any arms race stability anymore.
We need to introduce another concept here which is the idea that countries can depend upon deterrence in a world in which they cannot mount a defense. It’s useful to keep these words separate, defense and deterrence. Usually only wonky academics or people actually in the nuclear strategy business are careful not to confuse defense and deterrence. It’s useful to say defense is the ability to deny an enemy access to your territory and deterrence is the promise of punishment to discourage attack. Defense is therefore physical and deterrence is purely psychological. It’s useful to keep these two separate. Both of them can give you security.
And when it comes to strategic nuclear weapons, we rely on deterrence. Most Americans, Russians and Chinese don’t understand that they are defenseless in this world, and have been for decades. No country can mount a defense against modern ballistic missiles, never mind the new systems that are now being developed and deployed. So deterrence is our only hope. Once you admit that, it means that your ability to retaliate must survive an attack from an enemy; you’ve got to hold on to that concept. It’s sometimes called a “secure second-strike capability”. It confuses people, but that’s the term of art – secure second-strike. Since deterrence is punishment, you must preserve the capability to punish, no matter how hard you are hit and no matter how much the strike is aimed at your retaliatory capability.
So it’s the threat of punishment.
It is the psychological impact of the threat of punishment. If people don’t like this, they are very right not to like it. It’s not likeable to know that you can’t stop someone from hurting you – that you can only punish them afterwards, even if you’re dead. It isn’t particularly satisfying. But the argument of strategists has been that if you can be very convincing that the punishment will come, that the retaliation will come no matter what they do to you, there’s no “rational” reason for attacking in the first place. There’s no calculation that makes a first strike make sense.
And I would suggest that that’s what we have been assuming provided us with security. Here’s something else that troubles people: the value proposition of deterrence is unprovable. The notion is that if you did not have your deterrent, you would suffer an attack and therefore you say deterrence worked.
Well, this is a counterfactual. The logician will tell you it’s a proposition whose antecedent is known to be false. The proposition is that if you didn’t have the deterrent, you would have been attacked, but you did have the deterrent. And while historians and political scientists love counterfactuals, the logician will tell you “Sorry, you can’t say deterrence worked. You can’t know that. You can only say deterrence failed (if it does) because you will know if there’s a nuclear strike on your country”. So, you know that there was no war between the Soviet Union and the United States for 50 years and you strongly suspect deterrence had something to do with it.
If I were in the United States Navy, I would tell you that our undersea deterrent has guaranteed the peace. It may be true, but since it was always there you can’t say that it was the cause of the absence of war, a major war between the superpowers.
Back to arms control: if you don’t have arms control you’re not going to achieve security from deterrence. Deterrence really depends upon the ability to manage the competition between states. That’s one of the things arms control does for us. Let me give you a concrete example. If you just let the Strategic Rocket Forces of the Soviet Union or Curtis LeMay and the American strategic nuclear folks run rampant, they would build more stuff. As they built stuff, they would want to have the capability to limit damage to their own country, as well as to destroy or retaliate against another country, and they would build stuff that would attack the other fellow’s forces.
And the other fellow will do the same thing to us and pretty quickly we both look across the span of oceans and continents and say “They’re planning a first strike to disarm us so we will not have a secure second-strike. We won’t have retaliatory capability”. And so counter-force – the idea that you attack the other country’s forces first before you destroy his people and society – that’s what needs to be managed. Because if you leave retaliatory capability in place, it’s reasonable to assume that it has given you security by the fear of retaliation, but if someone starts targeting your capability to retaliate you start thinking, “Well, they would like to be free from the fear of retaliation and deny us our security”. That can happen in two ways – it can happen by building overwhelming offensive capability that can destroy the other fellow’s retaliatory capability and it can also happen – and here’s the big one – by deploying defense.
Now this really drives people crazy. The person it really drove crazy was Ronald Reagan.
There is a story told by more than one person, that when Gorbachev and Reagan were meeting in Reykjavik, towards the end of their meeting Reagan apparently said something like “And we will continue to build our defensive capability (known widely by the derisive term of Star Wars) and Gorbachev said “No, no, of course, you can’t. The whole idea of reducing our forces depends on you not building defenses and Reagan is reported to have said, “Mikhail, come on! It’s defense and you know our defensive systems can’t attack you”. Gorbachev responded “Don’t be ridiculous – your defensive systems deny us retaliatory capability and that’s what our security depends upon. Your defense is offensive”.
He did not say that US defenses destroyed crisis stability, but they do, and they destroy arms race stability. There are certain kinds of offensive capabilities that are particularly destructive of deterrence and stability, and a certain kind of defense also.
We have spent 30–40 years trying to build a strategic defense of our homeland and there’s a fear among our potential adversaries that we will succeed, which is provoking an arms race and crisis instability. So how do you control all this? I don’t know of any way other than arms control.
What about the argument that submarines offer a secure second-strike retaliatory capability and everybody’s got them so there should be no problems?
There are a few problems with that solution. One, as far as we know, at least US submarines are essentially undetectable and therefore survivable and therefore provide a secure second-strike. Here’s the problem: we’ve worked for decades to develop an anti-submarine capability and we have attack submarines that don’t have ballistic missiles aboard them that have an essential mission for the United States: to attempt to track Soviet boomers (now Russian) – ballistic missile carrying submarines – so that they will not be available in the event of a crisis. If you remember The Hunt for the Red October book, there’s a lot in there about anti-submarine warfare that had never been de facto declassified before and we all learned a lot about our US efforts to track Soviet submarines coming out of port and going to what are called their bastions.
So I would say that there is a concern that anti-submarine warfare will succeed at some point, especially given the development of persistent monitoring. This could have an impact on all attempts at securing your second-strike through mobility. Both the Chinese and Russians have mobile land forces. We do it through undersea mobility; they do it on land too.
There has always been a concern about ASW breakthroughs. Persistent monitoring, for example, with drones, could pose a threat to any system whose survivability has been based on mobility.
The second concern is about communication. Submarines at sea will survive; those in port won’t survive. The question is command and control of those submarines at sea – will that deter the other side? If the other side believes it can break that connection and has a long time to track down submarines, there’s a problem.
Third – this is one that will bend your mind a little bit – it’s possible that a first disarming strike has gone after ICBMs, bombers and submarine pens but not targeted cities or “countervalue” targets. The question for the surviving leadership in the United States is “what is to be gained at that point by using the remaining retaliatory capability at sea against a country which has preserved some of its offensive capabilities and is holding our cities hostage?” In other words, would we be deterred from retaliation by a well-conceived offensive attack in which only our submarines survived? Would warfighting make sense then? If you had other capabilities, then the other side might think, no this doesn’t make any sense; all we’re going to do is get into a tit-for-tat war. But if all we had left were submarines, that’s not so good.
Submarines are the backbone of our deterrent. Former Secretary of Defense Bill Perry, one of the smartest guys on the subject, has written more than once, most recently in his book The Button, that he would do away with our land-based ICBMs because they’re just sitting there as targets. ICBMs are obviously first-strike weapons from the perspective of the other side; they’re not retaliatory weapons because they wouldn’t survive. Perry said we don’t need that leg of the triad and he would invest more heavily, I think, in our undersea deterrent, making it more dynamic and survivable. And he would certainly not have us continue to pursue defense because that’s the other shoe here that has to drop: it is not just offensive systems that are destabilizing, that demand controls, but defenses also. In a way, defense is more pernicious than anything. When we deployed THAAD with its radar in Northeast Asia, we told the Chinese, this is not aimed at you, this is aimed at the North Koreans. And we told the Russians, about our BMD deployment in the Atlantic, this is not aimed at you, this is aimed at Iran’s future ballistic missile capability.
Neither party got any solace from our assurances. Despite what we said, the Russians and Chinese were thinking of the effectiveness of their retaliatory capabilities and perceived anything we did in their neighborhoods with defense as attempting to deny them their ability to achieve security through deterrence.
Defenses as Provocative
Defense is a very provocative concept and that’s why the Russians have tried so hard to get defense into arms control talks and not just talk about offensive systems, as we have in the past. You know, getting out of the ABM Treaty has left us unconstrained in what we can do in defense and I think that’s going to be a problem.
What defensive capabilities could we possibly trade away that would assuage the Chinese and the Russians?
Well, I don’t want to craft our arms control position now, but I would say that the objective would be exactly opposite of our objective in the development program that we have been pursuing with various technologies for decades. We would want to craft language that would prohibit us from continental defense. The issue is not really the AEGIS system, or THAAD or the Patriot system. It is our effort to defend the continental United States. A President would have to look the American people in the eye and say “My arms control objective here is to achieve our security and the only way we can do that is to not have the Russians and the Chinese constantly trying to build more offense”. Remember Putin’s speech to the National Assembly in March 2018 where he described a series of new nuclear systems with charts and videos? There was a torpedo with a thermonuclear warhead to destroy American coastal cities and there was a cruise missile not only nuclear-armed but nuclear-powered that would be able to attack from any vector, and so would defeat our homeland missile defense system. There was a hypersonic glide vehicle that would be launched by a very heavy ballistic missile and it would come off that ballistic missile and could attack from multiple directions also at very low altitude. In short, there was nothing we could shoot down with any current concept of strategic defense. These new systems were designed to defeat a defense which we haven’t deployed but are threatening to deploy. BMD advocates in the United States will tell you about all the new technologies and suggest that we will be able to defend the American homeland. As long as that is our dream, and I know a lot of people want that to be our dream, then the other side, because of the character of deterrence, is going to have to figure out ways of penetrating our defense because their security depends not on building a defense like ours, it depends on their ability to retaliate after our attack on their forces. And our defense would deny them that.
There is another concept from arms control that I want to introduce right now. It’s called the offense-defense cost exchange ratio. As long as the offense-defense cost exchange ratio looks like it does, where an increment of offense is cheaper than an increment of defense, an arms race will proceed and the defense will always lose – each defensive deployment overcome by a cheaper offensive deployment. You’ll have arms race instability and, ultimately, crisis instability.
So, with the latest Chinese test of a hypersonic glide vehicle, the appropriate response is not to say “We have to build up our hypersonic glide vehicles.“ The appropriate response is to say “let’s talk about controlling them.“
Let’s talk about arms control and why you need a hypersonic glide vehicle.
I mean we don’t want the other side to be able to launch a disarming first strike against us. A defense that is aimed at defending our retaliatory capability is not, I repeat, not destabilizing. When we did the SALT I treaty and we tacked on the ABM Treaty, allowing defense at first one site and then two sites to be built, the Russians defended Moscow.
But we didn’t defend Washington or New York or Los Angeles, we defended our ICBMs in North Dakota. The idea was to preserve our retaliatory capability. That’s not threatening to the other side. Our submarines aren’t threatening because they’re so survivable that they are considered a retaliatory system, not a first-strike system. That gives both sides a sense of security. It’s just that the Russians are doing things now to overcome a US missile defense which doesn’t really exist. Continental Defense doesn’t really exist. I mean, we have a certain number of defensive missiles and if the other side would tell us when they’re going to attack, describe exactly the trajectories and promise not to use any decoys or too many incoming missiles and certainly not any MIRVed warheads, well we’ll be in good shape about 50% of the time. That’s not a very impressive capability. That’s why I wouldn’t say we have continental defense. But the other side worries about it. So, if we don’t like these new Russian and Chinese systems, we’re going to have to reassure our potential adversaries that they don’t need to deploy these new systems to have survivable second-strike capability.
And by the way, while I’m on the subject of things that we do that really diminish our security rather than enhance it, the idea that we will eventually deploy Conventional Prompt Global Strike is another problem. The Chinese and Russians view our enthusiasm for being able to strike any spot on the globe with overwhelming force within one hour of the decision to do so as threatening their deterrent. We’re not deploying normal conventional systems, which means we’re using what are generally considered to be strategic systems for conventional missions. Our potential adversaries fear that our pursuit of this global strike capability is an attempt to deploy weapons that would take out their retaliatory capability without using nuclear weapons. The Chinese have been very specific that if we attack their strategic systems with conventional systems, they will regard that as they would the use of nuclear weapons against them and respond appropriately. Now, why would they say that? Because they’re looking at Conventional Prompt Global Strike and they’re not amused.
That seems like a tough one for the United States to walk back from.
Yes and no. We have quietly taken warheads from some missiles on some submarines and replaced them with smaller-yield nuclear weapons. Another bit of ideology here is that credible deterrence requires us to have the ability to retaliate against an enemy strike with a weapon whose yield is comparable to the attacker’s weapon and not one of higher level yield, lest we diminish the credibility of our threatened response. If we lose credibility we’re not as effective in deterrence. If they plan to use a small yield weapon – they being particularly Russia in a European context of some kind – we may not deter them because we’re not credible threatening to respond with a 100 kiloton-yield weapon if they have used a 2-kiloton weapon. There was a general argument in the Trump Administration’s Nuclear Posture Review (NPR) for the expansion in the number of our low-yield nuclear weapons in order to deter their low-yield nuclear weapons. It was this kind of thinking that was behind the decision to just pick a warhead on a missile in one of the tubes in a few of our submarines and replace it with a smaller one.
But this is ridiculous because U.S. warheads were already dial-a-yield.
The exact flexibility of the dial-a-yield weapons is not clear. We don’t want any country thinking that they can use low-yield weapons delivered by short range systems with impunity or that we would not regard those as sufficient provocation to retaliate with nuclear weapons of our own because it would be disproportionate. We don’t want them making that calculation. We can have a declaratory policy, we can be clear that we have the capability of varying our yields. I’m not sure that we need to match their yields exactly. I think we want to establish a real clear concept of a firebreak between conventional and nuclear weapons. We would regard any use of a nuclear weapon as the initiation of a nuclear strike without regard to the magnitude of the yield. That could be worked into somebody’s wonky speech somewhere and they would pick that up very quickly. We should preserve the firebreak as enormously important.
Yes, I agree. It almost feels like we’re moving backwards to a time you know, in the 1960s, when we had many thousands of very low-yield nuclear weapons, because we were afraid In Europe that we couldn’t deter a conventional strike.
Exactly. This is the effect of an asymmetric conventional force balance, which leads the side on the short end to try to compensate. The first book by Henry Kissinger on the subject provided the intellectual basis for tactical nuclear weapons, but a decade later, McNamara reversed all that and Kissinger revised his view on the desirability of tactical nuclear weapons.
How to Bring Stability to Cyber and Space Realms
What roles do cyber and space capabilities play in strategic stability and arms control?
We need to put our heads together on determining what is most dangerous about cyber and space and look for an arms control solution – a negotiated solution. Can we think of something that can be monitored? We need to look for breakout capabilities. The first thing we want is to have an understanding about not attacking satellites that provide security by warning, satellites that warn of an impending attack, or attacks that would blind the other side, where it must respond to defend itself against something that may not be happening.
As we have noticed in the past, transparency can be the friend of security, because it can be reassuring and we probably need to negotiate the preservation of transparency. I mentioned ASATs just now, but the same thing could be true about cyber-attacks. There are options in addition to prohibition, such as cooperative measures. We need to be careful about balancing what we are doing. For example, we may not want to totally foreclose offensive cyber capabilities. A cyber-attack that avoids kinetic action can achieve some useful objectives without lethality. But we have to look carefully at the strategic implications of leaving those options open.
The arms control challenge is to manage a situation through transparency and verification procedures that will improve everyone’s sense of security. Some ASAT may be manageable and not destabilizing. The problem will be, of course, if a country’s conventional defense depends upon blocking satellite capability that’s integral to the other side’s operations. For example, how do the Chinese think about their ASATs and our dependence on satellites for communication, intelligence and other functions in time of conventional war?
Of course, the ability to put satellites in orbit has spread to more countries. In the first instance an agreement between the US and the Russians would be valuable. I think we’re getting to the point where including the Chinese would also be useful, as hard as it is politically and practically. But I wouldn’t hold up US-Russian arms control by making an agreement contingent on China joining. Eventually, I think inclusion of the Chinese would be appropriate.
How does strategic arms control or strategic stability affect regional security balances, if at all?
Is arms control relevant to the Taiwan situation? To North Korea’s uranium program? To the Iranian program? To stability on South Asia? Potentially.
To me it isn’t a situation in which there are obvious technical things to be done, as there are at the strategic level. Those crises are not driven by technological breakthroughs. With the Taiwan contingency, this is not the sum total of our dispute with China. If we imagined that Taiwan was lost, had become like Hong Kong, would we be finished with US-Chinese confrontation? I don’t think so. Until China truly becomes a status quo power, we’re going to be off to the races with them over something. Taiwan is a worst case, for us, both politically and in military terms.
Disclosure Statement
No potential conflict of interest was reported by the author(s).
Additional information
Notes on contributors
Sharon Squassoni
Sharon Squassoni is Research Professor of the Practice of International Affairs at the Elliott School of International Affairs, George Washington University. After a career in the US Arms Control and Disarmament Agency, State Department and Congressional Research Service of the Library of Congress, she has focused her research agenda on practical ways to reduce risks from nuclear energy, fissile material and nuclear weapons.