Note:

Below is the text of remarks Dr. Ford presented on February 12, 2013, at the U.S. Department of State’s Foreign Service Institute (FSI).

Good afternoon everyone. It’s a pleasure to be back at FSI.   Your instructors asked me to come in today to offer my perspective upon arms control.   In light of what I remember the prevailing ethos to be at the State Department regarding arms control issues, and in the interest of challenging you a little bit, what I’d like to do is say a few words to problematize and “de-theologize” arms control for you in intellectually useful ways.  I’m certainly in no way opposed to arms control in principle, but neither am I committed to it in principle. As a public policy choice, I think uncritically embracing arms control measures is as foolish as uncritically opposing them.

Arms control is simply one among many possible tools that national leaders can use in order to serve other ends – national security, strategic stability, international peace, competitive advantage, or whatever – and, like most other tools, it’s probably wrong to assign it intrinsic merit or demerit. It can be helpful or unhelpful; it can be used well or unwisely. Details matter.

Arms control should be judged on the basis of what one does with it: it should be pursued where it would be useful and avoided where it wouldn’t.  When stated that way, of course, this sounds like no more than basic common sense, but you might be surprised how often the notion of arms control – like “disarmament” itself – is reified and valorized on the basis merely of a priori assumptions of its per se value.  Let me therefore state clearly at the outset that I think arms control has no a priori value.  Life is a lot more complicated than that.

I.            Types of Arms Control

Before I discuss some of these complexities, let me start by offering a typology of arms control measures.   I break arms control down into three categories.

1) First come bilateral and multilateral agreements and arrangements related to limiting, reducing, proscribing, and/or dismantling some sort of weaponry or other military-related technology.  I call these capability - regulatory measures. This is most intuitively obvious form: it is the traditional bread-and-butter of the arms control world.   It aims to control what capabilities one possesses.

2) Second are efforts to develop and promote “best practices” or codes of conduct pertaining to the use of certain types of technology or capability.  These I call behaviorally - regulatory measures: they relate not to what you have but what you can do with it.

3) Finally, there are steps related to transparency and confidence-building – that is, information-concessive measures. Into this bundle fall most transparency and confidence-building measures (T/CBMs).   These are often treated as a sort of auxiliary element in arms control, or as useful primarily with regard to facilitating verification of capability-regulatory measures, but I think they have real value in their own right and deserve to be considered a distinct form of their own.

One way or the other, these categories address themselves to the nature or scope of the threat states seem to pose to each other through their actual or potential possession of a particular type of military capabilities.  Indeed, I think one can conceive of all arms control as being most fundamentally about information – specifically, about trying to reduce the uncertainty that faces a country’s strategic planners. Capability-regulatory measures aim to reassure them that certain threats will not arise because potential adversaries will not acquire, or will restrain, a particular capability.   Behavioral measures seek to reduce the uncertainty that surrounds the employment of whatever capabilities exist, while information measures simply seek to make the parties better informed about the situation they face, whatever it is.

These three categories are not mutually exclusive forms of arms control, of course, and they may be, and often are, employed in combination, hopefully reinforcing each other.   One way or the other, however, they aspire to give planners more confidence about the nature of the future security environment, narrowing the landscape of alternative scenarios against which planning will prudently need to be undertaken – ideally by reducing the likelihood of more extreme threat scenarios, but at the very least by reducing the risk of unfavorable strategic surprise.

II.            Potential Benefits of Arms Control

So let’s think about the ways in which arms control can be a good policy choice. First, arms control can be used to slow or stop the course of a destabilizing arms race.  Arms races are not per se destabilizing, I think, but they certainly can be, especially to the extent that one side possesses some technological, economic, or other advantage that may enable it to pull ahead or “win” the race over time.   Even if an arms race does not necessarily intrinsically destabilize, moreover, it may be well worth impeding an upward spiral in order to permit the competitors to direct resources to worthier causes, or to prevent there from being such a weight of armaments that the results would be catastrophic if conflict does occur.   Capability-focused arms control can play a role in slowing or stopping such a race by setting limits upon what the parties are permitted to possess.

Capability-regulatory arms control can also be beneficial where it prevents possession of or reliance upon a particular type of capability in ways, or to a degree, that would tend to destabilize the relationship between the parties and make conflict more likely.   To give an example, many strategic analysts believe the deployment of ballistic missiles with multiple, independently-targeted re-entry vehicles (MIRVs) tend to be destabilizing because they make it easier for small numbers of missiles to take out large numbers of enemy systems, thus arguably encouraging preemption by an aggressor and/or launch-on-warning decisions by their potential victims.   The second Strategic Arms Limitation Treaty (START) of the 1990s would have prohibited MIRVing, and though the Russians refused to ratify it, it stands as an example of how arms control can be used in an attempt to keep parties from possessing not just more than a specific number of system but more specifically certain types of capability.

Arms control can also play a role in restraining parties in a reciprocally asymmetrical way – that is, across competitive domains – that conduces to stability, the avoidance of war, or other goals.   If one country most fears the other’s aircraft and that country most fears the first country’s ships, for example, it might be fruitless to try to negotiate aerial or naval capability restrictions alone (and especially “equal” ones); it might even be destabilizing if the parties did.  On the other hand, an arms control agreement that in some way limited aircraft while also constraining ships for both sides, would seem not just to offer something to both parties but actually to address the potential instability challenge presented by the specter of each side’s unconstrained pursuit of some special comparative advantage.

Multilateral capability-regulatory arms control of the sort that seeks to keep certain capabilities from proliferating beyond a pre-established group of possessors – as is the case with the Nuclear Nonproliferation Treaty (NPT), the most well known example of this type – can also play an important role in keeping things from spiraling out of control as the arrival of additional players creates an unmanageably complex and breakdown-prone matrix of deterrent balances.

Information-focused approaches “do” less than capability-restriction regimes, in the sense that they do not in themselves oblige any change in a party’s force posture.   Nevertheless, information-concessive arms control can sometimes provide real benefits by increasing parties’ understanding of the realities they face.   Naturally, the impact of such understanding will depend upon what this situation actually is, but T/CBMs can make a relationship more stable where they help dispel distrust and suspicion rooted in false perceptions that otherwise might spur the sides to adopt policies or acquire capabilities that could destabilize the balance between them.

The degree to which T/CBMs can ameliorate concerns about another party’s intentions is limited, since intentions are notoriously difficult to “know” with real assurance, and can in any event change.  Nevertheless, such measures can provide at least some window upon intentions, and can offer insight into another country’s ability to act upon its intentions, even if they prove malign.   Of course, the degree to which genuine transparency reduces distrust and fear will depend upon what is revealed.  It might be, for instance, that information-concessive measures serve not to dispel dark apprehensions but rather to confirm them.  (Sometimes there really are threats out there.)  All the same, even where transparency reveals threats, dispelling uncertainty and misconceptions about the lack of a threat – that is, eliminating one’s false sense of security – is hardly without value: no one wants to face unfavorable strategic surprise.

Arms control – or at least the posture of seeking arms control – can also have value as a diplomatic gambit, sincere or otherwise.  Some experts argue, for instance, that the mere pursuit of arms control agreements can have value, such as by making the symbolic point that one seeks peace and does not desire strategic advantage over a potential adversary.   (In this way, the seeking can itself function as a sort of confidence-building measure.   Perhaps not for nothing, then, does Article VI of the NPT oblige parties to “pursue negotiations in good faith” on nuclear disarmament; this may not be quite the empty gesture it is sometimes depicted as being.)  It may be, furthermore, that negotiating is sometimes useful as a way of making a point to third parties , whether or not agreement is achieved, or even expected.  Talks can sometimes serve to show the public in your own country or elsewhere, for example, that you are interested in a peaceful solution, even if the other side is not.   Negotiations can also be used to show that one has “gone the extra mile” before resorting to other measures.

(This is not to suggest, of course, that any of these indirect effects would outweigh the costs of accepting a bad agreement.  Nor is it to say that the posture of mere “seeking” will necessarily produce any benefits at all in any particular case, much less any benefit that would outweigh the potential cost of validating foolish principles even if an actual agreement were clearly not forthcoming.  These are empirical questions, not matters discernible  a priori.  Nonetheless, one must admit at least the possibility of such value.)

It may also be that reaching an arms control treaty may have value for one or more parties to some extent irrespective of the agreement’s actual content – as perhaps with Russia in the late 1990s and early 2000s, when it often seemed that getting some kind of legally-binding strategic treaty with the United States was most prized simply as the symbolic coinage of Moscow’s continued status as a superpower.   Sometimes, indeed, the act of negotiating may itself amount to giving the other side a boon or a concession – as seems to be the case today with North Korea, which shows not the slightest sign of being willing to give up its nuclear weapons (and in fact seems to have tested a new one not long ago) but nonetheless seeks nuclear talks with the United States in order to feed its own self-image as a nuclear weapons state and a world power that others must take seriously.

So what about behaviorally-regulatory arms control?  Such behavioral approaches seek to channel participants’ policy choices away from forms of competition or interaction that are particularly dangerous or provocative, and even though such controls are necessarily less constraining than capability controls – for it is easier to change behavior than to develop new weapons systems – they can provide benefits to the extent that they succeed in doing so.

Not all behavioral promises are equally credible, of course. (Perhaps the classic problem case is the idea of a nuclear weapons “no first use” [NFU] rule, which it is very hard to imagine a nuclear weapons possessor actually following if compliance meant the likelihood of it losing a major conventional war.)  That said, they can still sometimes be useful. In this regard, I’m particularly fond of measures such as nonproliferation “best practices” such as the Missile Technology Control Regime (MTCR) and the Hague Code of Conduct on missile transfers.   These don’t really bind parties to do much, and probably won’t do much where it matters most, but they generally appear to elicit at least some greater degree of restraint that would otherwise have been the case.

As a final comment about when arms control can be useful, it’s also worth mentioning that capability-based arms control (in particular) can be used to negotiate deals that constrain a potential adversary’s area of comparative advantage while leaving one’s own essentially untouched.  This doesn’t occur only where a deal is expressly lopsided – e.g. , by allowing one side a dramatically stronger force posture – but also where an agreement treats the parties equally on its own terms while yet having an asymmetric net impact.   (This might occur, for instance, if a treaty prohibited a type of weapon in the development of which one side had distinct advantages, while leaving the other side free to compete in areas in which it enjoyed greater strength and the first party was weaker.)   Precisely because they tend to provide asymmetric advantages, such deals are not necessarily conducive to strategic stability, but it’s easy to see how they might be attractive to those who can get such favorable terms.

III.            Possible Pitfalls of Arms Control

The possibility of such asymmetry in net impact, however, leads us to the other side of the leger: the possibility that under certain circumstances, arms control can be harmful.  This would presumably be the case from the perspective of the other side in a lopsided deal, and indeed it has been the ambition of many an arms control proposal over the years to create net asymmetry for purposes of competitive advantage.   (The Soviet “nuclear freeze” advocacy of the early 1980s and the Sino-Russian “weapons in outer space” efforts of more recent years are classic cases.  The former was designed both to freeze in place a Soviet theater nuclear advantage and to preclude nuclear responses to Warsaw Pact numerical superiority in conventional forces.   The latter proposals have been designed to preclude possible future U.S. deployment of the space-based weaponry feared by Beijing and Moscow while leaving untouched their ability to threaten U.S. space assets with terrestrially-based systems.)

Arms control agreements, of course, can also create problems where they fix one party in place while the other cheats.  Verification doesn’t always have to be perfect, of course, for depending upon the circumstances, not all violations have huge significance.   But often they do, and if an agreement under which the opposing party can achieve such advantages – either without being detected at all, or at least quickly enough that the other cannot respond in time – then the seeds have been sown for disaster.

But there can be problems even beyond the problems of potentially undetected or unanswerable noncompliance.   More generally, precisely because it is the ambition of arms control – or at least capability-focused arms control – to constrain specified types of arms competition, potential pitfalls can also lurk in its very success in doing this. After all, strategic relationships are not static balances, but instead dynamic ones. The world changes over time, one’s security environment evolves, and the requirements of stability, security, and deterrence therein are not fixed. As a rule, in complex dynamic systems inflexibility is a recipe for failure: systems survive to the extent that they can maintain themselves at some “sweet spot” between chaos and rigidity, being able to maintain coherence and a sort of net stability over time, but not in such an ossified form that they are unable to absorb the inevitable perturbations that they encounter when interacting with their environment.

What security, stability, and deterrence require today is not necessarily what will be required tomorrow, and precisely to the extent that they do try to fix in place a particular force posture balance or technological mix, there is a danger that arms control measures can outlive whatever contributions to peace and security they might initially create. Should circumstances change over time, arms control might end up being destabilizing , by retarding one or more parties’ ability to adapt safely to new developments.

The possibility of such “lock in” problems is often implicitly recognized in arms control agreements themselves, which commonly contain not only withdrawal clauses but also “sunset” provisions pursuant to which the entire deal will eventually expire and thereupon be subject to renegotiation in light of the then-prevailing circumstances. The Anti-Ballistic Missile (ABM) Treaty of 1972 is a classic example of an arrangement designed for one state of affairs that one of its parties later decided had become maladaptive in the face of subsequent circumstances – in this case, not as a result of anything the other party had done, but because of the emergence of third-party threats – and from which that party thus withdrew.

Strategic arms control negotiators sometimes expressly leave room for the parties to an agreement to adjust their force mix as they see fit within a broad set of overall parameters. Such an approach – employed in the Moscow Treaty of 2002, and to some extent also in the so-called “New START” deal of 2010 – reflects the understanding not only that treaty parties do not face identical situations, but also that that their needs can evolve even during the duration of an agreement. As a result, it can thus be harmful to specify too much in an arms control agreement.

Flexibility within the terms of a treaty – that is, a willingness not to provide for complete control of parties’ future decision-making on things such as weapons acquisition and deployment – thus seems to be valued, presumably at least in part because it reduces the danger that capability lock-in will imperil the interests of either side.   Such measures are important responses to the risk of maladaptive “lock-in,” but the challenge is a more general one: even with such partial safeguards, the decision to engage in any arms control deal will need to balance the benefits to be had from a specified type of constraint against the potential costs of such rigidity.

Another potential pitfall lies in “displacement effects” – that is, the risk that constraining competition in one arena will result in its displacement to other arenas.   This is not always bad, of course.  If there is to be competition, it is perhaps better to have it occur in some ways than in others, and if arms control can constrain competitive dynamics in the most dangerous areas then it may be well worth its costs and risks. Outcomes in complex adaptive systems are hard to predict, however, and it may be that displacement to another arena turns out to be equally (or even more ) problematic even than the original competition.

(The Washington Naval Treaties of the 1920s that limited traditional capital ships, for instance, may have accelerated the great powers’ movement into new fields such as submarines and carrier-based aviation, which would turn out to be much more important tools of warfare – as well as providing Germany with an asymmetric advantage in the Atlantic in the early stages of World War Two and Japan with the tools for its infamous surprise attack of 1941.   Similarly, by constraining competition in missile numbers, the first Strategic Arms Limitation Treaty [SALT] between the United States and the Soviet Union may have accelerated and deepened the parties’ destabilizing reliance upon MIRVed weaponry.   And it is possible that in the unlikely event that the Comprehensive Test Ban Treaty [CTBT] ever comes into force, it will displace nuclear competition into areas such as the proliferation and replication of “pre-tested” designs, or the development of “gun-type” weaponry that doesn’t need testing in the first place, with uncertain implications for net global stability.)

Finally, one should remember that arms control has an opportunity cost. National leaders only have so much time, energy, and political capital available that can be “invested” in the pursuit of policy objectives.   As a result, seeking an arms control agreement necessarily means – as a practical matter – that one doesn’t work so hard on other initiatives.   This is fine if there aren’t more important things that need attention, but if one’s strategic circumstances are such that there isn’t much danger of a spiraling arms competition – or if one doesn’t face a dangerous adversary whose weapon acquisition or behavior one needs to constrain or at least better understand for one’s own strategic planning – it might be wiser to spend one’s finite reserves of political and diplomatic capital on other things.  Arms control should not be pursued out of mere inertia, or solely out of some politically-correct conception of virtue ethics that prizes seeking arms control for its own sake even where achieving it would contribute little or nothing to peace and security.   It should be pursued where the advantages to be had outweigh what one could do with that time, energy, and political capital on other policy fronts, but not otherwise.

IV.            Conclusion

On the whole, I think that arms control’s record is historically mixed and theoretically problematic enough to teach us that it is unwise to approach these matters without caution and an open-eyed awareness of arms control’s various potential risks as well as its potential benefits.

If anything, on the whole, I suspect that the conventional wisdom in much of the policy community – and in institutions such as the State Department, if in present company I might be so bold – rather overvalues arms control. Historically, arms control has sometimes been useful in helping manage aspects of strategic competitions that might otherwise have been more dangerous. It has not, however, been a mechanism capable of effecting significant change for the better in such strategic relationships. (It can codify and build upon change that occurs in the nature of strategic relationships, such as with the end of the Cold War, but the greatest successes of arms control have followed rather than driven strategic change.)   And arms control has real pitfalls, which constitute potential traps for the unwary.

Yet arms control can indeed have real value, and as I suggested at the outset of these remarks, we would be no less foolish to reject it on principle than to insist upon it on principle.   Details matter, complexity is pervasive, and I think it would be entirely unwarranted to posit a per se answer – either positive or negative – about the merits of arms control. Instead, the key lesson is to avoid ideological complacency, remembering that arms control is neither inherently bad nor inherently good. It is simply a tool, and if one wishes to use it well, there are many variables to take into consideration.  Arms control theory needs to be “de-theologized,” I submit, if arms control is to be practiced well.

Thank you. I look forward to our discussions.

-- Christopher Ford