Good day, and thank you for inviting me.
Every review conference of a treaty is, by definition, a chance to look back and review how countries have handled the challenges the instrument was designed to address, and how they have (or have not) lived up to its aspirations. It is also a chance to draw lessons from that history about how best to approach what challenges remain before us. The upcoming 50th Anniversary of the Nuclear Nonproliferation Treaty (NPT)’s entry into force is an especially good opportunity for such stock-taking — and for assessing our own way ahead — as we remember its successes, assess its limitations, and plan for the future by recommitting ourselves to the Treaty and to ensuring that it does at least as well in its second half century as it did in its first.
And I do think it’s important to look back on the NPT’s history, not least in order to remind ourselves of what we have accomplished — and thus what might be imperiled if we allow the nonproliferation regime to weaken. As I’ve pointed out many times, it is critical on this anniversary to remember the security benefits that the NPT has brought to the world by helping forestall the catastrophic proliferation of nuclear weapons and risk of nuclear war that experts feared in the early 1960s. It is also critical to remind ourselves of the ways in which the nonproliferation regime has provided a foundation for other benefits we all hold dear, such as by making possible worldwide sharing of the benefits of peaceful nuclear technology, and by helping keep proliferation from precluding movement toward disarmament. It is of paramount importance that we remember all this on the Treaty’s Golden Anniversary.
Seriousness and Unseriousness
But what I’d like to do today is take a slightly different approach, by exploring with you less what has been accomplished by the nonproliferation regime as a whole than drawing lessons from how some of this was done. I think such historical lessons can illustrate the complexities of living in a nuclear-armed world, as well as provide us examples to draw upon today from how thoughtful statesmen have wrestled with these challenges in the past.
In approaching the weighty issues raised by nuclear weapons, some people tend to treat issues of risk reduction and disarmament as if policymaking were no more than a stark morality play. In some quarters, the assertion of moral clarity on nuclear policy comes across as melodrama in which the goals seem to be to score rhetorical points, express outrage against those who disagree, and signal one’s own righteousness.
Now, one can get a lot of mileage out of such posturing from an NGO fundraising perspective, I suppose, and some states certainly seem to like using such moralism in efforts to demonize adversaries. But I would submit to you that such ways of thinking and speaking about the challenges of nuclear weaponry are powerful indicia of unseriousness. Rather than actually signaling virtue, they signal that the speaker either does not understand the challenges of nuclear policymaking in the first place, or understands them but glosses over the complexities in order to manipulate the listener. Neither alternative is particularly creditable.
To be sure, moral commitment and intensity are critical ingredients of good and principled policymaking. We need them, for example, to keep us on a sound compass bearing toward the right objectives, even — or perhaps especially — when the immediate terrain ahead is uncertain and there are obstacles in our path. We also need these qualities to help us have resilience in the face of setbacks, and to keep us plugging away at things when the going is hard and heavy. And we need them to help us keep checking in with ourselves, from time to time, to make sure that we still have confidence in our direction of travel.
My point is merely that there can be peril in overdoing such intensity. In excess, moral emotion in policymaking makes one less credible and less effective, and can lead to grave errors. It impedes the self-awareness we need to have as fallible humans trying to make tough decisions on the basis of uncertain information and unpredictable outcomes; it can breed fixation, possibly even fanaticism. One needs enough moral commitment to keep trying to do the right thing even in the most trying of times, but it must be tempered by enough thoughtfulness and intellectual humility that we do not allow rigidity and obsession actually to keep us from achieving our objectives.
In truth, therefore, there is not a tension but rather a complementarity between having a strong moral drive and being alive to nuance, complexity, ambiguity, and structural tension. Finding the right balance in this respect is no compromise. To the contrary, it is essential to success. Such balance is not important despite having morally compelling objectives; it is vital because they are so compelling. Serious policymaking isn’t identity-political performance art. When it comes to making real policy decisions — rather than simply impressing one’s social media followers or political constituents, or delivering propaganda points at a conference — serving a good cause badly is not our job; we need to get these things right in actual practice, precisely because these matters are surpassingly important.
So why am I waxing philosophical about such things in the context of the NPT and its history? Because nuclear policymaking does not reward approaching it with the blunderbuss subtlety of a campy melodrama; to the contrary, it tends to punish such moralism rather severely.
Studying our collective history in the nuclear realm can help us learn how to approach our collective future more wisely. Among other things, aspects of the NPT’s history can help illustrate at least two points to assist us in being prepared for its next 50 years. Familiarity with this history can provide a corrective for moralistic absolutism and oversimplification, by demonstrating how complex the various legitimate equities actually involved in nuclear policymaking can be. And it can provide insights into how such equities indeed can sometimes be balanced in responsible and enduring ways.
Such lessons are, I would submit, extremely important to all of us today. Whether we acknowledge it or not, those of us in government charged with such policy questions bear heavy responsibilities for stewardship of our countries’ national interests — and those of international peace and security more broadly — that I believe we would betray by cynical fluidity and moralistic obsession alike. It is our job to navigate the ship of state between these rocks, and the study of history can help us do so better.
Responsible Nuclear Stewardship: A Case Study
To that end, my case study in nuclear negotiating and the challenges of responsible nuclear weapons stewardship derives from the dilemmas that the United States faced in trying to maintain NATO’s nuclear deterrent against the threat of Soviet aggression while bringing to closure its negotiations with the Soviet Union over the core provisions of the international convention that would ultimately become the NPT. This example is valuable, I would argue, for a couple of reasons.
For one thing, it helps insulate us in the present day against propagandistic manipulation, for if you believe current Russian narratives, everything I am about to tell you is false and NATO nuclear policy violates the NPT. We should remember history in order to repudiate such disingenuous silliness. More importantly, however, this case study is important because it helps illustrate how responsible nuclear stewards struggle — but can nonetheless succeed — in balancing complex, interrelated equities. So let’s take a look.
As the NPT negotiations advanced, Washington found itself having to manage a number of important equities. All of these were legitimate and important, but not all of them seemed to point in the same direction, at least not easily.
One obvious one was the importance in the mid-1960s of bringing the NPT negotiations to a close. U.S. intelligence estimates had for some years been giving grim forecasts of the nuclear weapons proliferation that could happen as various countries around the world acquired the technical capacity to develop such weapons. Accordingly, there was a strong sense of urgency in successfully concluding a treaty that would help reduce these dangers. Ever since the so-called “Irish Resolution” at the United Nations in 1961, the conceptual core of the proposed treaty was the need for provisions both barring non-possessors from getting (or trying to get) nuclear weapons and barring possessors from giving non-possessors access to such devices or know-how. These twinned prohibitions indeed ultimately became Articles I and II of the NPT, but as of the mid-1960s there was not agreement between Washington and Moscow upon the details.
One source of such disagreement was the relationship between these core nonproliferation rules and another important equity the United States also badly needed to protect — specifically, the need to ensure the continued efficacy of the nuclear deterrent with which Washington confronted the Soviet Union as NATO and the Warsaw Pact faced off in Central Europe. With the Alliance numerically outgunned on the central front, it was critical to U.S. and NATO planners that they maintain a nuclear deterrent posture sufficient to deter the Soviet aggression that they most feared, and which they felt unequipped to meet with conventional arms alone. This meant that NATO needed to make its threats of nuclear retaliation against Soviet attack highly believable.
A third equity was the importance of constraining nuclear proliferation pressures within NATO, and this was greatly complicated by game-theoretical problems of deterrence within an alliance framework. Specifically, the challenge was to make nuclear deterrence credible not merely to Moscow, but indeed to the NATO allies themselves.
The construct of U.S. “extended nuclear deterrence” in support of NATO relied heavily upon Washington’s willingness to use its own nuclear forces to back up the Alliance, and U.S. planners went to great lengths to make this threat as credible as possible. Nevertheless, the fact remained that the anticipated ground battle for Europe would take place thousands of miles from the U.S. homeland, and would most immediately decide the fate of territories and peoples that were not, at the end of the day, American.
This led some Europeans to worry about the possibility of the Alliance being “decoupled” from Washington, inasmuch as a Soviet invasion that only threatened Europe would confront the Americans with the choice — in phrasings one might have heard at the time — of whether to sacrifice Chicago in order to save Hamburg. It was central to U.S. planning to make American nuclear weapons available to prevent defeat at the hands of Warsaw Pact armor in Europe, but some Europeans worried that, in a pinch, Washington might opt to avoid risking Chicago. Might the United States, they feared, be tempted to not intervene with U.S. nuclear weapons against a Soviet invasion if doing so would imperil U.S. cities?
At least some such concerns might be imagined to lie behind the British and French decisions to maintain their own, independent nuclear deterrents — but such thinking was not confined to London and Paris. West Germany, in particular, also had such fears, and much of NATO planning during this period was devoted to trying to provide an answer to this reassurance problem that did not involve the Germans acquiring their own nuclear weapons — as indeed some in Bonn did feel might be needed. At this point in the mid-1960s, Nazism and World War II were still very recent memories, and the thought of German fingers on the proverbial nuclear “button” was a worrying one even in the West, and a terrifying one in Moscow.
And here’s where it got particularly tricky, however, because some of the ways one might reassure allies against “decoupling” might be very problematic from the perspective of crisis stability vis-à-vis the Warsaw Pact. How could the Federal Republic of Germany (FRG) be made sufficiently comfortable with the reliability of nuclear deterrence — and thus dissuaded from itself engaging in weaponization — without steps being taken that might exacerbate the escalation-management challenges associated with having more Western fingers on more triggers, and which could potentially also provoke overreactions from Moscow?
When my story begins, NATO had already begun to consult with NATO allies about potentially making U.S. nuclear weapons available to some of them in time of war. But at this point in the mid-1960s, NATO was entertaining a further possibility: a Multilateral [Nuclear] Force (MLF) of jointly-owned and -controlled nuclear missile submarines manned by crews drawn from the various NATO nations.
The MLF, however, was opposed by Moscow with a vehemence that threatened, at the very least, to sink the global effort then underway to establish a nonproliferation regime. Worse still, these dilemmas threatened to make the deterrent standoff in Central Europe more unstable. What ability the MLF might give NATO allies to launch nuclear weapons independently was not entirely clear — at least not to Moscow, at any rate — and it might in any event give these allies access to some non-trivial amount of nuclear weapons knowledge. To the degree that Soviet planners imagined the FRG being able to launch NATO nuclear weapons against them at will, these discussions raised important questions of crisis stability vis-à-vis the Kremlin.
These challenges with the FRG thus wrapped together a complex brew of deterrence, reassurance, nonproliferation, and crisis stability issues. All of these equities came together in U.S.-Soviet negotiations over what would become Article I of the NPT: the Treaty’s provision prohibiting helping others acquire nuclear weapons.
The resolution of all this balancing is well known to anyone who has followed these issues. It occurred in three more or less simultaneous moves:
- First, the United States dropped the MLF concept (as well as the alternative Atlantic Nuclear Force concept) in favor of sticking with a construct of purely wartime nuclear sharing arrangements in which the Americans forward-deployed nuclear warheads in Europe that were slated for employment in conflict by Allied air forces, but in which the United States retained absolute control over these devices in peacetime and prior to any U.S. decision to implement any such wartime allocation;
- Second, the Soviets grudgingly accepted that this NATO posture was better than the available alternatives — and that it met Moscow’s fundamental security needs vis-à-vis potential additional NATO states’ access to nuclear weaponry — and they agreed to the text of what became the NPT, while conceding that NATO’s mere nuclear consultations and training for wartime operations did not present Article I problems; and
- Third, the FRG accepted that these arrangements met Bonn’s deterrence needs without the necessity of either MLF-style shared launch authority or indigenous weaponization.
This historic compromise successfully balanced the complex and simultaneously compelling equities of deterrence, reassurance, nonproliferation, and crisis stability — and this bargain has held ever since, at least so far. It is in this compromise that the extraordinarily valuable nonproliferation norms of the NPT and U.S. nuclear deterrence policy ended up working together to help establish, and provide a strong foundation for, the nonproliferation regime.
The Story in Documents
While its broad contours are well known, however, one new element in recent years is that thanks to the declassification of U.S. records from the years of the NPT’s negotiation, it is now possible to document this story in great detail. I will spare you today the blow-by-blow details of how U.S., Soviet, German, and other Allied officials worked out this momentous understanding, but I will put a more complete description in the longer version of these remarks we post on the ISN Bureau website. Last year the United States released a declassified negotiating history and NATO released a collection of contemporaneous documents, and we are working to declassify more of the original record.
For present purposes, one can treat this documentary story as beginning in October 1965, when U.S. National Security Council staffer Spurgeon Keeny wrote a memorandum to his boss, President Lyndon Johnson’s National Security Advisor McGeorge Bundy, observing that the draft NPT text that had been tabled by the Soviet Union would prohibit the proposed NATO MLF concept, and potentially any other NATO bilateral arrangements. This was obviously a problem for Keeny, since those arrangements represented U.S. policy at the time. Nevertheless, he also observed, presciently, that Moscow “might eventually propose to give up the language outlawing our NATO arrangements if we were prepared to give up the MLF.” Ambassador Jacob Beam, who headed the Arms Control and Disarmament Agency (ACDA)’s International Relations Bureau, made a similar point to Secretary of State Dean Rusk a month later, recounting that discussions with the Soviets had “conveyed an impression of a degree of possible flexibility on the subject of existing [nuclear] arrangements in NATO.”
After mulling over these ideas, President Lyndon Johnson wrote a letter to the Chairman of the Soviet Council of Ministers, Alexei Kosygin, in January 1996, declaring that the United States was “not prepared to enter into any agreement that would deny our allies the possibility of participating in their own defense through arrangements that would not constitute proliferation.” He also made pointedly clear, however, that Washington understood “proliferation” to occur only when “a non-nuclear nation acquires its own national capability or the right or ability to fire nuclear weapons without the explicit concurrent decision of an existing nuclear nation.”
With this letter to Prime Minister Kosygin — the version I have is undated, but it was probably written in January of 1966 — President Johnson thus clearly flagged to the Soviets the idea that notwithstanding prior MLF planning, NATO’s nuclear arrangements would stop short of allowing European allies possession or control of nuclear weapons: the weapons would remain U.S. weapons and under U.S. control, and Washington would retain absolute veto rights over their actual use. The United States exhorted the Soviets to accept this principle as the basis of their work on the NPT’s core prohibitions upon proliferation.
A few months later, in June of 1966, Spurgeon Keeny wrote another memorandum recounting that ACDA Director William Foster had talked about this with Soviet ambassador Anatoly Dobrynin — and that the Russians seemed willing to accept existing NATO nuclear arrangements (of strict unilateral, peacetime U.S. control and an American veto upon Allied employment) as long as “the Soviets could be sure that Germany would not use or would not be able to use nuclear weapons on their own decision.” This was echoed in August by a memo from Foster himself to Secretary Rusk, arguing that joint ownership of nuclear weapons would need to be ruled out, but that in return for this concession “our present bilateral arrangements within NATO” could likely be preserved. Foster made just such a case to President Johnson himself in September.
But an implied signal from Dobrynin was one thing, and actually getting Soviet agreement to an NPT text was quite another. Nevertheless, progress was made, and declassified memoranda from the succeeding few months show the two sides gradually circling in on that conclusion from their respective positions. To this end, Rusk met repeatedly with Soviet Foreign Minister Andrei Gromyko at the United Nations in New York. In these discussions — memorialized in U.S. memoranda — the Soviets were keen to ensure that the draft treaty would bar “transfers” of nuclear weapons either to individual non-weapon states or “through an alliance,” while Rusk reassured Gromyko that some such agreement seemed feasible.
The superpowers’ two top diplomats also explicitly discussed the fact that NATO’s arrangements, in which the United States maintained complete control over its forward-deployed weapons, only applied in peacetime: should a war actually break out, Rusk made clear, things would be different. This did not bother Gromyko, however, who replied that such wartime questions were mere “political considerations” rather than issues of legality under the draft NPT. (Neither side seemed to think it was either desirable or feasible to worry about such treaty provisions in the event of all-out war between two alliances that between them then possessed perhaps 40,000 or 50,000 nuclear weapons.) To help reassure Moscow that U.S. transfers of control or launch authority over nuclear weapons to NATO allies would never occur in peacetime, ACDA Director Foster met with Soviet Ambassador Alexei Roshchin later in September 1966 to make these points once more.
A U.S.-Soviet Working Group convened to work through the drafting challenges associated with these basic concepts was meeting by late September, and it reported back to the White House on September 30 that its members had agreed upon “compromise language for a non-proliferation treaty that both sides can live with.” This language would be “very nearly coextensive with present U.S. nuclear policy,” in that it would permit the continuation of the NATO status quo in which “U.S. nuclear weapons available for use by allied forces assigned to NATO in the event of hostilities could … be transferred to those forces in that event.” The Soviets had agreed to drop draft treaty language upon which they had previously insisted, leaving only text that acceptable to the Americans since it would not apply to “consultative and planning arrangements of the type contemplated within NATO.” To make sure there was no misunderstanding, Rusk met with Gromyko again in early October to reaffirm their understanding that the treaty should focus on what was prohibited, not what was permitted, and to reinforce the message that the treaty would not ban U.S. nuclear consultations with NATO allies — that is, the kind of training and contingency planning that had then already been formalized within the Alliance.
All that really remained, then, was to convince the Germans. Accordingly, ACDA Director Foster reached out in January 1967 to the FRG’s ambassador in Washington, Karl Heinrich Knappstein, to recount that the Soviets seemed to have become reconciled to NATO’s existing consultative arrangements, had dropped language that would have prohibited “training of allied troops for possible use of nuclear weapons in the event of war,” and seemed to understand that delivery vehicles (such as allied dual-capable aircraft [DCAs]) would not be covered by the treaty as long as no actual “transfer of warheads or control over them” occurred. This didn’t mean that the Soviets wouldn’t criticize NATO’s arrangements, Foster stressed, but he made clear to the Germans that Moscow had indeed abandoned its prior argument that those arrangements would be unlawful under the draft treaty. National Security Advisor Walt Rostow later made a similar point to German parliamentary leader Rainer Barzel at the White House in February 1968: the Soviets, Rostow made clear, now “know they cannot raise the [NATO nuclear-use] double-key question or the question of nuclear consultation” as NPT problems.
These understandings with the Soviets and the Germans clinched the deal making clear that NATO’s “nuclear sharing” arrangements were not a problem under Articles I and II of the NPT. Under Secretary of State Nicholas Katzenbach duly summed things up in a letter to Secretary of Defense Clark Clifford in April 1968, recounting that as a result of the Rusk-Gromyko negotiations in 1966, the NPT’s Article I now protects NATO “alliance consultations on nuclear defense” and “nuclear defense deployment arrangements.” An attachment to Katzenbach’s letter explained that the NPT would permit transfers of delivery vehicles or delivery systems “so long as such transfer does not involve [nuclear] bombs or warheads.” It also stressed that if “a decision were made to go to war … the treaty would no longer be controlling.”
The various sides had thus reached a compromise that neatly handled all of the complex tensions and interrelationships between the important equities involved. The Soviets killed off the MLF, the Americans gained agreement that NATO arrangements were not barred by the NPT, and Germany retained the reassurance provided by an Alliance system that promised it the ability to deliver U.S. nuclear weapons against Warsaw Pact targets in wartime. It had proven possible to reach agreement on a nonproliferation treaty; NATO could still rely upon its existing arrangements to ensure Alliance “coupling” in the interests of deterrence; Germany and other NATO allies felt reassured enough about their likely degree of active involvement in the event of nuclear war not to pursue nuclear weapons on their own; and the Soviets did not have to worry about a new peacetime status quo in which Germans had their fingers upon the proverbial atomic trigger. Not too bad, I’d say.
So thus was history made. Notably — except for recent attempts by Putin regime propaganda to pretend, in effect, that none of this negotiating ever happened, and therefore to convince historically ignorant listeners that NATO’s nuclear policy is somehow in violation of the NPT — this historic compromise has lasted to the present day. It still remains critical to NATO’s nuclear deterrent concept.
We should remember this compromise as we approach the 50th Anniversary of the NPT’s entry into force. It can help remind us that actual policymaking in the nuclear arena requires the management and balancing of multiple legitimate but partly competing equities — and is thus entirely unlike the simplistic morality play that some would have it seem.
The story of negotiating the final text of the NPT’s Articles I and II illustrates how difficult it is to have a nuclear posture, and to conduct diplomacy, in ways that maximize the deterrence of aggression, minimize escalation risk and crisis instability, maximize ally reassurance, and minimize proliferation pressures — all at the same time. It also provides a case study in responsible nuclear stewardship, visible in how the Treaty’s negotiators worked through the ways in which these interrelated challenges then manifested themselves.
In the fraught security environment of our own time, we face our own challenges — not least with regard to how to do all of these sorts of things and to explore constructive ways forward that reduce tensions and strengthen trust (as the NPT exhorts us) in order to facilitate nuclear disarmament. Informed and inspired by the lessons of our past, and keenly aware of the complexities of nuclear policymaking and the responsibilities of nuclear stewardship, we are working to do this in multiple ways.
First and foremost, of course, we are continuing to modernize our own nuclear forces to avoid the block obsolescence of our strategic delivery systems and thereby preserve the nuclear deterrence upon which our security — and that of our friends and allies — has depended for many decades. It is important to bear in mind, however, that all “modernization” programs are not the same. For our part in the United States, we are replacing like with like — i.e., simply replacing older systems with newer versions of the same thing, and in comparable numbers: a new ICBM, a new strategic ballistic submarine, and a new manned bomber.
These U.S. plans stand in sharp contrast with Russia’s push to add to its strategic arsenal with exotic new systems, such as the “flying Chernobyl” of its accident-prone nuclear-powered cruise missile and a nuclear-powered underwater drone. Our modernization also stands in sharp contrast with Russia’s anticipated expansion of its number of non-strategic weapons, and with China’s perilous track to at least double the size of its nuclear arsenal — while also expanding the range and diversity of its delivery systems — over the next ten years.
We are also modernizing the nuclear deterrence aspects of our NATO relationships in order to keep this longstanding capability viable in the years ahead, even while continuing to adapt the Alliance to the modern challenges of conventionally-armed deterrence forced upon us by Russia’s aggression against its neighbors and belligerent posturing against us and our allies. We hope that our NATO allies will remain true to this path, and live up to their commitments in these regards, for our collective security depends upon it.
At the same time, we are doubling down on diplomacy in order to help build the foundations for a better security environment in the future. We are, for instance, implementing a new multilateral disarmament dialogue, based around our Creating an Environment for Nuclear Disarmament (CEND) initiative, that reconceptualizes and reframes global disarmament discourse to redirect it from its past fixation upon mere manifestations of underlying problems in the strategic environment to a serious exploration of how to start addressing the problems themselves.
This dialogue, while still in its early stages, continues to show promise. Sixty-two participants from 31 countries, including the United States, met at Wilton Park in the United Kingdom on November 20-22 for the second meeting of the CEND Working Group. Participants attended from NPT nuclear-weapons States and non-nuclear-weapons States alike, as well as from some nuclear weapons possessors who are not signatories of the NPT at all. These Participants began to lay the groundwork for translating the CEND dialogue into action by developing Concept Notes for each of the three CEND subgroups on:
- Reducing perceived incentives for states to retain, acquire, or increase their holdings of nuclear weapons and increasing incentives to reduce and eliminate nuclear weapons.
- Mechanisms to bolster nonproliferation efforts and build confidence in and further advance nuclear disarmament.
- Interim measures to reduce the risks associated with nuclear weapons.
The next meeting of the CEND Working Group will take place early next year.
We are also developing new approaches to arms control that aim to include both Russia and China, for the first time, in an accord to prevent a destabilizing new global arms race and help build a more stable security environment. And we are helping lead the global charge to develop norms of responsible behavior and “best practice” standards in new security domains being created by complex, rapidly-evolving, and potentially transformative technologies — such as in cyberspace and outer space — that can intersect with traditional deterrence approaches, but which in their very complexity are resistant to traditional legal-prohibitive arms control.
There is nothing simple or easy about these steps, nor about the complicated relationships between them. Doing all this is difficult — and none of what we are about is by any means guaranteed to succeed. But that is how things work in the real world; struggling to preserve and advance such interwoven equities in a complex environment is what responsible players do.
I hope that the upcoming anniversary of the NPT’s entry into force can remind us of the importance of wrestling with such problems responsibly today — and in the future — as our predecessors did in bequeathing to us that important Treaty and the global nonproliferation regime of which it today forms the cornerstone.