Bilateral strategic stability: What the United States should discuss with Russia. And China.

By Robert J. Goldston | October 14, 2021

In the first US “salvo” test of ground-based missile interceptors, two interceptors were launched from Vandenberg Air Force Base in California on March 25, 2019. They successfully intercepted a “threat-representative” ICBM target launched from a test site on Kwajalein Atoll in the Marshall Islands. This photo shows the launch of the “lead” interceptor, which destroyed the missile’s reentry vehicle. The “trail” interceptor struck the remaining “most lethal object” it could find. Credit: Missile Defense Agency In the first US “salvo” test of ground-based missile interceptors, two interceptors were launched from Vandenberg Air Force Base in California on March 25, 2019. Credit: Missile Defense Agency

At their mid-June summit, Presidents Biden and Putin agreed that “the United States and Russia will embark together on an integrated bilateral Strategic Stability Dialogue in the near future that will be deliberate and robust. Through this Dialogue, we seek to lay the groundwork for future arms control and risk reduction measures.” Deputy Secretary of State Wendy Sherman, who headed the negotiations that led to the Joint Comprehensive Plan of Action with Iran, is leading this activity from the US side. The second meeting with the Russians took place on September 30, in Geneva, involving both initial substantive discussion and the organization of working groups.

There are at least four topics that should be on the table in the dialogue with Russia:

  • Limits on ballistic missile defense
  • Elimination of silo-based intercontinental ballistic missiles
  • Prohibition of intermediate-range nuclear-armed missiles
  • A common nuclear declaratory policy

And recent events suggest the need for bilateral strategic stability dialogue with China—on the same four topics.

Limits on ballistic missile defense. Paradoxically, US attempts at ballistic missile defense have made the United States—and the whole world—less safe. Russia is concerned that an extensive US exo-atmospheric ballistic missile defense system might ultimately be effective, so it is developing hypersonic delivery systems to skim warheads along the top of the atmosphere, long-range nuclear-powered torpedoes to carry high-yield warheads into ports, and—most bizarrely—long-range nuclear-powered cruise missiles to attack the United States over the South Pole. It is also building massive, liquid-fueled, silo-based, intercontinental ballistic missiles (ICBMs) capable of carrying both many nuclear warheads and many “penetration aids,” to overwhelm ballistic missile defenses in a more conventional manner. Sadly, the ballistic missile defense system Russia is working so hard to evade or overwhelm has failed repeatedly to perform successful intercepts, even in highly scripted tests.

So, while the US investment in ballistic missile defense motivates this panoply of new weapons, it provides no significant additional security against Russian attack. This is reminiscent of how President Reagan’s attachment to ballistic missile defense, aka the Strategic Defense Initiative (SDI) or Star Wars, derailed the opportunity at the Reykjavik summit in 1986 for a world-changing agreement with Secretary General Gorbachev on the elimination of nuclear weapons.

The possibility that ballistic missile defense might ultimately work raises the specific concern for Russia that the United States might be able to launch a powerful first strike on its nuclear forces, and then “mop up” the remaining Russian warheads headed for the US homeland through ballistic missile defense. This concern provides an incentive for Russia to strike first if it believes that the United States is about to strike, an example of extreme strategic instability.

A further problematic feature of ballistic missile defense: It stymies even moderate arms reductions. If both sides agree to reduce their warhead count, say by an equal amount, the side with stronger ballistic missile defense is advantaged. Even a reduction in the ability to mount penetration aids disadvantages the side with weaker ballistic missile defenses.

The United States is concerned about potential Iranian nuclear and missile capabilities, but Iran cannot develop anything approaching the Russian nuclear arsenal in the foreseeable future. It should be possible, therefore, to agree on fully verified limitations on US ballistic missile defense capabilities, as a necessary part of a deal to reduce the total number of deployed strategic weapons on both sides.

Elimination of silo-based intercontinental ballistic missiles. ICBMs located in fixed silos are sitting ducks. Unless they are set for launch under attack, they can be of little use for deterrence of a nuclear attack nor for war fighting once a nuclear exchange has begun. However, when set for launch under attack, they present a danger of mistaken launch either because of a false alarm—for example due to cyber-intrusion—or because of miscalculation in the fog of conventional or sub-strategic nuclear war.[1],[2],[3] They also all but force sole presidential authority for launch under attack as there would be no time for serious consultation with other officials. The 1981 Office of Technology Assessment report on MX Missile Basing[4] stated:

Basing MX missiles in silos and relying on launching the missiles before a Soviet attack could destroy them (launch under attack, or LUA) would be technically feasible, but it would create extreme requirements for availability of, and rapid decision-making by, National Command Authorities. … there would always remain the possibility of error; depending on the nature of the error, it couId mean a successful Soviet first strike against MX or it couId mean a nuclear war started by accident.

Silo-based missiles in a launch-under-attack posture are not survivable weapons, so they are much more effective for a first strike than for response to attack. If a US president judged that a first strike was likely to come from Russia, there would be a strong incentive for the United States to strike first. The same would be true of Russia, but even more so, because of the large number of warheads and penetration aids on Russia’s current and planned large silo-based missiles, making these silos particularly attractive targets. Since both sides know that the other side knows that striking first could be advantageous, the problem is compounded. This is, indeed, the classic example of strategic instability—the topic of the upcoming dialogue.

There are a number of options to eliminate silo-based ICBMs consistent with the five-year New START extension. New START sets an upper limit of 1,550 on the sum of nuclear warheads on intercontinental ballistic missiles (ICBMs), on submarine-launched ballistic missiles (SLBMs), and on the number of nuclear-armed bombers, which are counted as carrying just one warhead. Under New START and its five-year extension, either side can adjust the balance of its strategic nuclear arsenal and reduce its total count.

As shown in the table below, actions can be unilateral or bilateral. Each of these actions divides into an option in which the warhead count is reduced by the number associated with silo-based ICBMs, and an option in which the ICBM warhead count is transferred to submarines and strategic bombers.

The unilateral reduction option is simple: eliminate all US silo-based ICBMs and their associated 400 warheads. There would be no need to negotiate with Russia, and it could be accomplished as a Presidential Nuclear Initiative, such as was implemented by President George H.W. Bush at the end of the Cold War. The advantages of this option are that all US nuclear forces would be off of launch-under-attack posture, and the total number of warheads available to cause mass death would be reduced. A potential domestic political problem is that some would object to this option because the United States would have fewer deployed nuclear warheads than Russia, even though the remaining 1150 nuclear warheads could very effectively destroy Russian civilization (not to mention instigating a devasting nuclear winter for the whole globe).

Scenarios for the Elimination of Silo-based ICBMs

ICBM = intercontinental ballistic missile : SSBN = submarine armed with ballistic missiles : LUA = launch under attack : BMD = ballistic missile defense

The unilateral transfer option would mitigate the above political problem but would not reduce the total number of deployed warheads. However, it would fill some of the existing empty slots in US ballistic missile submarines and so partially reduce the strategic instability associated with “uploading” of additional warheads at a time of enhanced tension, which could motivate the other side to strike before the uploading is complete.

According to the Nuclear Information Project of the Federation of American Scientists,[5] the total number of SLBMs on the 12 US submarines at sea at any one time is currently 240, carrying approximately 860 nuclear warheads. Two other nuclear-powered submarines, carrying about 40 SLBMs, are typically in refueling overhaul. Up to eight nuclear warheads can be carried on each SLBM, and a total of about 1,895 such warheads are available for use. So the maximum total number of SLBM warheads at sea could be about 1,624, or 764 above the number currently deployed. As a result, there is ample room for the 400 count of warheads on existing silo-based ICBMs to be transferred to SLBMs. Alternatively, warhead counts transferred to new B-21 Raiders or older B-52s with Long Range Stand Off (LRSO) cruise missiles would add the stabilizing feature of delivery vehicle recallability. Neither ICBMs nor SLBMs can be recalled once launched.

RELATED: Reimagining the Open Skies Treaty: cooperative aerial monitoring

The bilateral reduction option takes away the need for a launch-under-attack posture for both the United States and Russia and reduces both deployed arsenals by about one quarter. Publicly available estimates[6],[7] suggest[8] that the number of Russian silo-based warheads is about equal to the US value of 400, so that the warhead-count reductions required to eliminate silo-based ICBMs would be similar on both sides. As discussed above, however, it should be anticipated that Russia would insist on verifiable limits on US ballistic missile defense capabilities, since reducing warhead and/or penetration aid count disadvantages the party with weaker ballistic missile defenses.

The final bilateral transfer option would see both the United States and Russia eliminating the need for a launch-under-attack posture, but it would not reduce the total arsenals on either side. Ballistic missile defense limitations would likely still be required, since the number of penetration aids that Russia could mount would likely be reduced. It would reduce both sides’ upload potential, a positive feature. In the case of Russia, as for the United States, slots would need to be found for about 400 warheads previously located on silo-based ICBMs. There should be about 200 unfilled slots on existing Russian road-mobile ICBM launchers, although SS-25 road-mobile missiles capable of carrying 274 to 455 warheads may be retired soon. On the other hand, upgrades and replacements are ongoing to Russia’s road-mobile ICBMs.

The Russian fleet of ballistic missile submarines is estimated by the Federation of American Scientists4 to carry 624 warheads in deployed boats, with some room for further upload. Russianforces.org5 estimates the total capability at 656 warheads. Although one of the Russian missile subs is likely to be retired soon,4,5 four new Russian submarines are expected to be added to the Russian fleet in the next few years, each capable of carrying up to 96 warheads. So there appears to be ample room for Russia to keep the full New START warhead count under the bilateral transfer option for the duration of New START.

In any of the above options, U.S. strategic bombers, whether new B-21 Raiders or older B-52s, could be placed on strip alert at the air bases where the ICBM silo fields had been located, providing the positive domestic political impact of replacing the local economic benefit of the silo-based missiles. Strip alert would render the bombers invulnerable even to a “bolt from the blue” attack by Russia, without requiring them to be launched on a non-recallable mission, as is the case for ICBMs. This would mitigate any perceived technical risk that US submarines in the deep ocean might someday become targetable.

It should be recognized, however, that even a single surviving US nuclear missile submarine, carrying up to 100 thermonuclear warheads in these scenarios, would devastate Russia. This reflects the Reagan/Gorbachev dictum, recently repeated by Biden and Putin, that a nuclear war can never be won.

Prohibition of intermediate-range nuclear-armed missiles in Europe. The United States pulled out of the Intermediate Range Nuclear Forces Treaty after Russia was found to be violating its terms. This treaty was broader than necessary[9] to address the needs of bilateral strategic stability, as it outlawed all US and Russian land-based intermediate-range missiles, whether nuclear or conventional, anywhere in the world. The issue of bilateral strategic stability, however, is tied to the presence of short-flight nuclear-armed missiles in Europe, including Russia west of the Urals, that could decapitate Russian or European leadership. In principle it should be straightforward, therefore, to agree that intermediate range nuclear missiles will not be located in Europe, by either side. Verification can make use of the same approach as the New START treaty, in which neutron detectors are used to verify that objects declared not to contain nuclear warheads do not emit neutrons. National technical means can be used to locate launchers.

Common nuclear declaratory policy. A valuable capstone for a successful bilateral dialogue on strategic stability would be a common, stabilizing statement of nuclear declaratory policy.

The most stabilizing nuclear declaratory policy would flow from the verified elimination of nuclear weapons, consistent with the Treaty on the Prohibition of Nuclear Weapons. If there are no nuclear weapons, there can be no nuclear war. This would be a welcome but highly unlikely outcome of bilateral dialogue with Russia. Russia considers itself to be confronted by superior conventional NATO forces and has made clear that it considers that nuclear forces may be necessary for its defense against conventional attack.

Current Russian policy[10] states that the “[t]he Russian Federation reserves the right to use nuclear weapons in response to the use of nuclear and other types of weapons of mass destruction against it and/or its allies, as well as in the event of aggression against the Russian Federation with the use of conventional weapons when the very existence of the state is in jeopardy.” The policy specifies the conditions under which it is possible that Russia would use nuclear weapons:

  • Arrival of reliable data on a launch of ballistic missiles attacking the territory of the Russian Federation and/or its allies.
  • Use of nuclear weapons or other types of weapons of mass destruction by an adversary against the Russian Federation and/or its allies.
  • Attack against critical governmental or military sites of the Russian Federation, disruption of which would undermine nuclear forces response actions.
  • Aggression against the Russian Federation with the use of conventional weapons when the very existence of the state is in jeopardy.

The second most stabilizing nuclear declaratory policy is one of “no first use.” If no one uses nuclear arms first, again there can be no nuclear war. But for the same reason as above this is not a likely outcome of bilateral dialogue with Russia. On the other hand, in parallel with the bilateral dialogue with Russia, the United States should initiate serious discussions with its allies and partners as to whether they would realistically want the United States to use nuclear weapons to defend them against conventional attack, rather than act solely as a deterrent to nuclear use against them. First nuclear use by the United States could subject these nations to devasting nuclear counterattack, likely far worse in consequence than a protracted conventional war, even one that began with an initial successful conventional military invasion. The confidence of our allies and partners in our support has been shaken in recent years and may take some time to first re-establish and then strengthen, but just the fact that this will be a difficult task is no justification for shirking it.

RELATED: Do Germany and the Netherlands want to say goodbye to US nuclear weapons?

The United States and Russia might be able to agree in the near term on a policy that nuclear weapons would only be used as a measure of “defensive last resort.” This concept was first put forward by McGeorge Bundy, William J. Crowe Jr. and Sidney Drell[11] in 1993, and has been re-iterated by this author[12] and by Perkovich and Vaddi[13] in closely related forms. In effect, this policy would state that, “We would only consider the use of nuclear weapons to defend our states, and our allies and partners, against imminent existential threat from nuclear-armed adversaries, and then only once all other means had been exhausted.” This is close to recent published US[14] and Russian nuclear declaratory policy noted above, but simpler and arguably clearer in intent and more restrictive. Surprisingly, in one way this is stronger than “no first use,” in that a nuclear response to a non-existential nuclear attack that could be countered by conventional means would not be considered.

Colin H. Kahl, President Biden’s under secretary of defense for policy, spoke on June 23 at the 2021 Carnegie International Nuclear Policy Conference. He said, in substance, about nuclear declaratory policy,

“I’m not going to tell you what our declaratory policy is and how much it will change, because ultimately that’s a decision the president will make, and the president hasn’t made that decision. … I think that we can all agree that nuclear weapons are a credible deterrent against existential threats. But I think people of big intellects and good faith can disagree about how explicit or ambiguous we should be about scenarios under which we might consider the use of nuclear weapons below that threshold. And that is a debate that we will have internally.”

This and the Russian posture make it clear that moving to a nuclear declaratory policy in which nuclear weapons would only be considered for use as a defensive last resort against an imminent existential threat from nuclear-armed adversaries would be a significant step for either side. But perhaps such a step would not be impossible to take—particularly on a bilateral basis. If this could be achieved, it would constitute a significant advance. 

What about China? In parallel with bilateral discussions with Russia, the United States should engage in bilateral strategic stability dialogue with China. Chinese concerns about US ballistic missile defense capabilities are legitimately even greater than those that Russia has, because the Chinese nuclear arsenal[15] is much smaller than the Russian. And since North Korea has proven nuclear explosive and ICBM capabilities, the United States legitimately has even greater need for ballistic missile defense capabilities in the Far East. This situation may be driving China to expand its nuclear forces in order to assure US vulnerability to Chinese retaliation, and so maintain strategic stability. But an arms race between the United States and China will be self-defeating for both sides. As with Russia, a clear understanding is needed, including reliable verification, that will assure China that US anti-ballistic missile capabilities will not enable the United States to undertake a disarming first strike against it.

Disturbingly, there are open-source intelligence reports[16],[17] that suggest China is building new ICBM silo-fields, with about 250 new silos under construction. It is likely that China’s warhead count will be limited by its available stockpile of fissile material.[18] Nonetheless, Chinese silo-based ICBMs are fundamentally destabilizing, in the same manner as US and Russian ones. The United States would be motivated to destroy these easily located silos early in a conflict, before they could be used; China would be motivated to use them before they could be destroyed.

The United States has a complex challenge to negotiate the number of silo-based ICBMs downward at the same time with both Russia and China. But the only other choice would allow the numbers of these destabilizing weapons to spiral upwards. Both China and Russia may be attracted to silo-based ICBMs because these are the most cost-effective means to be able to launch large numbers of warheads and penetration aids at the United States; such missiles can carry multiple independently targetable re-entry vehicles (or MIRVs), assuring deterrence in the face of effective ballistic missile defenses. Therefore, the only safe path forward is to constrain US anti-ballistic missile capabilities, in a fully verifiable manner, as regards both China and Russia.

Land-based intermediate-range nuclear-armed missiles are less of a consideration with China than with Russia, since there are no US-allied nuclear-armed states within intermediate range of China and no real prospect of locating land-based US nuclear missiles in East Asia. But discussions leading to prohibition of such missiles in Asia would certainly be stabilizing and would contribute to assurance of US allies in the Far East, helping to make a US “no first use” policy attainable.

This brings us to the question of a common nuclear declaratory policy. China has previously committed itself to “no first use,” but it is less explicit about its nuclear declaratory policy than the United States or Russia. The United States faces particular difficulty with its Asian allies and partners on the question of nuclear declaratory policy, since China is strengthening its conventional military capabilities. It may well be that, even with China, the most realistic common policy that can be achieved in the near term, in parallel with the discussions with Russia, is one of “defensive last resort.” Even if this is so, focused effort to provide sufficient assurance to our Asian allies and partners to allow a mutual declaration between the United States and China of a “no first use” policy is a critically important element along the path to ultimate global nuclear disarmament.

In sum. The world is at a fork in the road. One path forward leads to a safer world, potentially free of nuclear weapons, while the other leads to rapidly increasing risk of nuclear annhilation. Bilateral strategic stability dialogue is a promising means to choose the safer path. Success will require the participants in these dialogues to view themselves as sitting on the same side of the table, trying to solve together the problem of reducing the vast nuclear threat to all humanity.



[1] R. Goldston, “Why the new US ICBMs would be too expensive even if they were free,” Bulletin of the Atomic Scientists, January 3, 2020

[2] David Wright, Wlliam Hartung, and Lisbeth Gronland, “Rethinking Land-Based Nuclear Missiles,” Union of Concerned Scientists, June 2020

[3] Matt Korda, “Siloed Thinking: A Closer Look at the Ground-Based Strategic Deterrent,” Federation of American Scientists, March 16, 2021

[4] Office of Technology Assessment, “MX Missile Basing,” Library of Congress Catalog Card Number 81-600133, September 1981

[5] H.M. Kristensen and M. Korda, “United States nuclear weapons, 2021” Bulletin of the Atomic Scientists, 2021, Vol. 77, No. 1, 43-63

[6] H.M. Kristensen and M. Korda, “Russian nuclear weapons, 2021,” Bulletin of the Atomic Scientists, 2021, Vol. 77, No. 2, 90-108

[7], accessed on 6/10/2021

[8] Kristensen and Korda estimate that Russia currently has 604 “slots” available for warheads on silo-based missiles. ( estimates this number at 578.)  Kristensen and Korda also estimate the total warheads deployed on silo-based and mobile ICBMs at about 2/3 of the total number of available slots, suggesting that the number of Russian silo-based warheads is about equal to the U.S. value of 400.

[9] R. Goldston, “A short and practical take on the INF: Small (and focused) is beautiful,” Bulletin of the Atomic Scientists, June 28, 2019

[10] “Basic Principles of State Policy of the Russian Federation on Nuclear Deterrence” Ministry of Foreign Affairs of the Russian Federation, 8 June 2020,

[11] M. Bundy, W.J. Crowe Jr., and S. Drell, “Reducing Nuclear Danger,” Council on Foreign Relations, 1993, Vol. 72, No. 2, 140 – 155

[12] R. Goldston, “What can the Kerch Strait incident teach about nuclear declaratory policy?” Bulletin of the Atomic Scientists, December 20, 2018

[13] G. Perkovich and P. Vaddi, “Proportionate Deterrence: A Model Nuclear Posture Review,” Carnegie Endowment for International Peace, January 21, 2021.


[15] H.M. Kristensen and M. Korda, “Chinese nuclear forces, 2020,” Bulletin of the Atomic Scientists, 2020, Vol. 76, No. 6, 443-457

[16] J. Lewis, “China is Radically Expanding Its Nuclear Missile Silos,” June 30, 2021,

[17] M. Korda and H. Kristensen, “China Is Building A Second Nuclear Missile Silo Field,”

[18], accessed on August 3, 2021

As the coronavirus crisis shows, we need science now more than ever.

The Bulletin elevates expert voices above the noise. But as an independent, nonprofit media organization, our operations depend on the support of readers like you. Help us continue to deliver quality journalism that holds leaders accountable. Your support of our work at any level is important. In return, we promise our coverage will be understandable, influential, vigilant, solution-oriented, and fair-minded. Together we can make a difference.


Get alerts about this thread
Notify of
Newest Most Voted
Inline Feedbacks
View all comments
alice slater
alice slater
6 days ago

What a paltry proposal from the nuclear weapons establishment, now writing for the Bulletin of Atomic Scientists. With a new treaty signed and ratified for the Prohibition of Nuclear Weapons which finally banned the bomb and closed the legal gap in the old Non-Proliferation Treaty which merely called for “good faith” efforts for nuclear disarmament, asking now for certain missile controls, etc. is ridiculous. The world is facing catastrophic climate destruction and with 14,000 nuclear bombs on the planet, thousands of them primed and ready to shoot in minutes, and the US planning to spend over a trillion dollars in… Read more »

Last edited 6 days ago by alice slater
Robert Goldston
Robert Goldston
5 days ago
Reply to  alice slater

My reading is that BMD is the big hold-up in making progress on arms reduction with Russia and China, which is why I put it first on the list.