Good afternoon everyone, and thank you very much to Rusty Ingraham and the other organizers of this course for the chance to speak to you today. It's always a pleasure to be out here on the FSI .
I. The Modern Debate
I often find it fun to talk about arms control, because different parts of the U.S. policy community have very different instincts on the subject, giving arms control debates a very different character from nonproliferation discussions. We're all relatively consistent here in Washington on nonproliferation issues, with pretty much nobody thinking that more countries should acquire nuclear weaponry. We may disagree, inside the Beltway, about the details of how best to go about trying to ensure that, but despite divergences in atmospherics and rhetoric, there has been much nonproliferation policy continuity from the George W. Bush Administration into the Obama Administration.
But the story is different with respect to arms control, where at issue is how to approach regulating the world's existing arms. Here, community often splits into feuding camps, the members of which don't get along very well. They may succeed each other in office, but them seldom actually talk to each other.
Arms control tends to produce debates that are remarkably absolutist and, well, theological. If you'll permit me a bit of caricature, on one end of the stereotypical continuum is a hawkish cabal that opposes any restrictions upon U.S. power, and which thinks arms control is always bad – a trap, if you will, set by internationalist America-bashers to constrain our power while favoring that of our actual or potential adversaries.
On the other end is a gang of credulously uncritical arms control addicts, who are convinced that it is always good, unhappy with any American muscularity, and devoted to the reduction of stockpiles as a moral duty that must be fulfilled as quickly as possible, and irrespective of the strategic consequences. Each side thinks the other is at least stupid, and quite possibly actually malevolent.
This description is a cartoon, of course, and unfair to both sides. Nonetheless, there's probably enough truth in it to make my point: our public discourse on arms control is grievously polarized. For my part, I want to confound both sides of this absolutist debate by de-theologizing arms control policy, because I think both ends of this ideological continuum will make bad policy choices if left entirely to their own devices. Arms control issues are too important to be left to absolutists.
Having said that, we don't have all afternoon, so I need to limit my remarks. Make no mistake: I believe that arms control can contribute to U.S. , as well as to international peace and security. But since I suspect that none of you disagree with me on that – FSI, after all, is a State Department institution – it would be wasting your time to preach to the choir about the value of arms control.
Precisely because this is a State Department program, however – and we all know where State's instincts lie with regard to the potential for diplomatic negotiations and international treaties to solve all the world's ills – I'll focus my comments on what I hope will rattle you a bit. I want to talk about when arms control doesn't work so well, and about the traps that it can sometimes create for the unwary diplomat.
I wouldn't argue that arms control always creates such problems, of course, any more than I'd argue that it is a per se good. If you leave with one "take-away" lesson today, I want it to be that these are emphatically not questions that can be answered a priori. These are empirical matters, and the wise diplomat needs to be aware of arms control's potential to do both good and harm. He or she needs to be able to engineer good outcomes where this is possible, but also willing to pull the plug – and to walk unflinchingly away – when this cannot be done.
II. Critiques of Arms Control
Let me start by examining some potential ways in which arms control can be problematic simply by failing to do what it aspires to do. These possibilities are relatively well known, and not too surprising: (1) verification failure; (2) enforcement failure; and (3) manipulation of the process.
One obvious challenge is how to detect violations. This is a common locus of complaint in evaluating specific arms control measures. Opponents of the Comprehensive Test Ban Treaty (CTBT), for instance, worried a great deal about the possibility of clandestine violations, and this concern helped lead to the U.S. Senate's rejection of CTBT in 1999. With CTBT at least theoretically still on the Obama Administration's agenda, this issue remains salient today, since it is increasingly believed that Russia and perhaps China have been secretly conducting yield-producing nuclear tests as part of an ongoing program of developing new varieties of weapon, all undetected by the CTBT's monitoring organization. Proponents of any agreement will naturally be expected to answer questions about its verifiability, and about the potential impact of undetected violations.
But clandestine cheating is hardly the only potential worry. One must also be able to detect cheating in time to be able to do something useful about it. Reliable verification, after all, accomplishes little if it's just a tool for documenting one side's strategic defeat. Detection must be timely enough to permit an effective response. Timely warning is a challenge recently raised, for instance, in connection with the relationship between International Atomic Energy Agency (IAEA) inspection schedules and the "conversion times" needed to divert safeguarded capabilities to prohibited uses.
The effectiveness of the range of available responses, once detection occurs, is also a critical factor, for it would be small consolation to get lots of warning if one lacks an effective means to respond. Nor is a lack of tools the only potential problem. Verification and compliance policy can face political challenges even where tools exist, if detection of a violation would create the need to confront difficult and painful choices such as whether to withdraw, undertake a countervailing arms buildup, or go to war. In such circumstances, pressures can arise to look the other way and avoid having to acknowledge that cheating is taking place.
(This isn't just a curmudgeonly hypothetical, by the way. Versailles Treaty arms inspectors during the interwar years are said to have downplayed evidence of German rearmament, for example, because their home governments were unprepared to deal with it and did not want to know. President Clinton also once admitted that U.S. proliferation sanctions laws created incentives for his administration to "fudge" intelligence assessments in order to downplay proliferation transfers, lest acknowledging them disrupt diplomatic relationships such countries such as China.)
Not all arenas of arms competition are equally susceptible to effective verification and reliable enforcement, and it pays to think long and hard about such challenges in considering how to structure a deal – or whether to negotiate in the first place. Even if you choose to accept a weak deal, it's vital at least that this be understood: trouble awaits if you plan for your future security while mistaking a weak system for a strong one.
The savvy arms controller also needs to know that quite apart from whatever agreement is actually reached, arms control negotiating can itself be used for manipulative and destabilizing purposes. A state that wishes to develop a prohibited category of weaponry, for instance, might find it useful to go through the motions of negotiating over the issue in order to prevent or delay outsiders from undertaking countermeasures during the time it needs to finish developing that capability, or at least to hide or protect the associated infrastructure.
Such a strategy of negotiation in order to buy time for strategic positioning has an ancient pedigree. Before the outbreak of the Peloponnesian War between the ancient Greek city-states of Athens and Sparta, Athenian envoys drew out negotiations with Sparta over a mutual ban on defensive walls in order to win time in which their city rebuilt its own walls higher and stronger than ever. More recently, this is the sort of thing that Iran and North Korea seem to have done in negotiations over their respective nuclear weapons programs. (Iranian nuclear negotiator Hasan Rowhani publicly bragged about doing this in 2005, but the tactic apparently still works pretty well.)
These are very real pitfalls, and the wary diplomat needs to watch out for them. For the most part, however, these potential problems represent cases in which the difficulty arises from some incompleteness in the arms control process, where arms control has failed on its own terms. I thus find them less interesting, analytically, than instances in which arms control destabilizes because of its success in imposing constraints. So let's look at those possibilities now.
III. Arms Control and Stability
The conventional wisdom of the diplomatic community holds that arms control is essential for strategic stability, especially in a nuclear-armed world. And while arms control certainly can be valuable, let's stretch our cognitive envelope a little by thinking a bit about when that wouldn't be true.
When might the actual success of an arms control agreement potentially lead to instability? I will suggest four basic ways: (1) direct destabilization; (2) maladaptive "lock-in"; (3) competitive displacement; and (4) manipulative substance.
The first type, direct destabilization, is probably likely to result simply from a basic conceptual failure as one designs one's arms control agenda in the first place. To explain the kind of thing I mean, let's take as an example the most ambitious agenda for modern arms control: weapons elimination in the form of complete nuclear disarmament. If nuclear weapons are part of the reason that the great powers haven't gone to war against each other for many decades, then their successful abolition might destabilize, by removing one important reason for such war-dissuasive conclusions.
There is certainly no lack of historical examples of states that have prized nuclear capabilities for their presumed effect in deterring large-scale conventional military attack, among them those of the NATO states during the Cold War, and Pakistan, Israel, and Russia today. If there is anything to such assumptions, however, a world free of nuclear weapons might be more militarily unstable than before. As I once heard one foreign diplomat from a non-nuclear-weapon state say at the Conference on Disarmament, disarmament might "make the world safe again for large-scale conventional war."
One might perhaps argue about the degree to which this is true, but to me it is striking that so many disarmament activists do concede – at least privately – that any "nuclear zero" would need to be accompanied by some kind of psychopolitical "transformation" of international politics in order to ensure that a post-abolition world doesn't end up looking like 1914 or 1939. To my eye, if you need to resort to such magical thinking to make your strategic theory make sense, it probably doesn't. Either way, however, it should be clear enough that even successful arms control is not always immune to the law of unintended consequences.
Just because the abolition end-point suffers from such conceptual problems, of course, doesn't that all steps in that direction it necessarily do. They may, or they may not. Details matter, and my distaste for categorical, a priori answers in this field drives me to think that there can be no substitute for looking at proposed measures on their individual merits. My point here is merely to use the asymptotic case of abolition as an illustration of the fact that it is possible for arms control measures to be built on structurally defective foundations, and that it behooves us to approach arms control policy with intellectual humility – with an awareness that we are fallible creatures, and that it is not utterly beyond question that some of our most strongly-held strategic-theoretical certainties will turn out to be fallacious.
But what of my second category: "lock-in"? Since the world is a dynamic place, it may also be that even an arms control formula that is initially stabilizing – as it is intended to be – can turn out to be destabilizing over time. Rigid rules can sometimes create brittleness, and fixing in place a particular technological or numerical status quo may not always be a good idea.
Imagine, for instance, that future U.S. and Russian negotiators agree to cut their forces down to a strategic "monad" of the type of delivery system classical American nuclear theorizing regards as being most "survivable" and thus likely stabilizing: submarine-launched ballistic missiles (SLBMs). Into the resulting world, however, let us a few years later introduce a "wild card." The agreement, after all, would have increased both sides' incentives to search for new methods of strategic anti-submarine warfare, so let's imagine that one side gets clever, or lucky, and finds a good one. This would represent catastrophic strategic surprise for the other side, and in this case arms control would have set the stage for instability by tying one side to a single-point-of-failure force posture, upon which it would now be unable to rely in deterring potential aggression.
Arms control constraints may thus, over time, create problems by "locking in" a force posture unsuitable to changing circumstances. This is why agreements usually contain withdrawal clauses, and sometimes also "sunset" provisions that in effect prevent them from fixing anything in place for too long.
The history of the Anti-Ballistic Missile (ABM) Treaty illustrates these potential challenges. It was rooted in a theory of strategic stability pursuant to which missile defenses were deemed destabilizing because they would encourage a spiraling offense/defense arms competition.
What seemed like a good idea to the Americans in 1972, however, did not look so compelling in the post-Cold War era. Indeed, under President George W. Bush, American officials concluded, in effect, that the treaty had locked in a status quo that under modern circumstances was becoming destabilizing. After the collapse of the USSR and the end of the Cold War, the Russo-American arms race had been reversed, and both sides were dramatically reducing their arsenals. At the same time, the United States had come to perceive an emerging missile and nuclear threat from third parties such as North Korea and Iran. In this new context, the anti-defense status quo of the treaty came to be seen in Washington as maladaptive, fixing in place rules no longer necessary for their original purpose and that now seemed likely to empower rogue states to use their emerging arsenals to bully their neighbors or even threaten the great powers, who were largely defenseless against long-range ballistic missile attack.
The ABM Treaty, however, had no "sunset" provision; it was intended to prohibit defenses forever. Nevertheless, it had a withdrawal clause, and in December 2001 the United States announced its intention to withdraw. Today, the treaty is a dead letter, and even the Obama Administration's somewhat scaled-back missile defense plans involve capabilities notably beyond what the 1972 agreement would have permitted. One can argue back and forth about the merits of withdrawing, of course, and some still do. My point here is just to point out that the "lock-in" effects deliberately created by arms control may not always be stabilizing: even if you do get the recipe right at first, circumstances can change.
An third thing to watch out for is the possibility that arms control will displace military rivalry from one arena into another. Such displaced competition may not always be as bad as unregulated competition, of course. But there's no rule that says this has to be so; things might actually get worse. It could be argued, for example, that the numerical limits imposed by the Strategic Arms Limitation Treaty (SALT) of the early 1970s on U.S. and Soviet delivery systems helped push the superpowers more into the deployment of multiple, independently-targeted re-entry vehicles (MIRVs). (Unable to aim at more targets by building more missiles as they had previously done – but still wishing to be able to hit more targets – the powers invested in ways to do so with their existing missile force. MIRVs were that answer.)
But this was problematic, for nuclear analysts tend to believe that using MIRVs – at least on missiles in land-based silos – is less "stabilizing" than single-warhead missiles, because MIRVs seem to make it more attractive for an adversary to strike preemptively. (A single attacking weapon, hitting a MIRVed missile in its silo, can take several enemy warheads out of action, but this advantage is lost if the other side is permitted to fire first.) Since SALT encouraged the superpowers to move to pervasive MIRVing, it arguably left the arms race more "unstable" than the negotiators found it. The savvy arms controller should worry about such displacement effects: be careful your "cure" doesn't end up being worse than the disease.
A final trap to watch out for is good, old-fashioned strategic manipulation – that is, deliberately skewed substantive outcomes – in the form of deals the other side seeks precisely because their effect would be destabilizing in its favor. Think, for instance, of Soviet support in the early 1980s for a "nuclear freeze" and the idea of "no first use," which clearly aimed to impede NATO nuclear responses to Soviet deployments of new ballistic missiles and to the Warsaw Pact's numerical superiority in conventional forces. Similarly, Russian and Chinese proposals for a treaty to prevent an arms race "in Outer Space" have long been phrased to preclude what they feel a potential future U.S. advantage in space-based weaponry, while leaving untouched those countries' own ability to threaten our space assets with their terrestrially-basedanti-satellite capabilities. Not all arms control is good arms control, and getting "no deal" can be better than one that the other side has "cooked." (If you happen to "cook" a deal for your own side, of course, I suppose someone will probably give you a nice medal and a promotion. But watch out for the other guy, and think hard about what he wants you to accept.)
IV. Conclusion
Anyway, that's my laundry list of potential arms control pitfalls of which the wary diplomat needs to be aware. Do these landmines along the road mean that arms control is invariably a bad idea, and that it cannot contribute in valuable ways to U.S. security and international peace and security? Of course not. It is often very valuable indeed, and in the unlikely event that you doubt this, we can talk about it in the question-and-answer period if you like.
But while remaining committed to getting real value out of arms control wherever possible, arms controllers also need to be smart, historically and contextually aware, attuned to the multiple ways in which good intentions can go wrong, and willing – in the right circumstances – not just to say "no" but to say "hell, no." This is really just common sense, but in today's theologized arms control discourse, it sometimes needs to be said anyway. Traps await the unwary if our enthusiasms outrun our wisdom.