The Discrimination Problem: Why Putting Low-Yield Nuclear Weapons on Submarines Is So Dangerous


The United States has the most diverse and potent nuclear force on the planet, capable of deterring and, if necessary, defeating and destroying any military and any nation on earth. The Trump administration’s recently released Nuclear Posture Review doesn’t think that’s enough. Going beyond the modernization program that upgrades and maintains the existing force, the document calls for a variety of capabilities and missions for American nuclear forces that have long been on Republicans’ wish list. Specifically, the document places a renewed emphasis on expanding the role and size of the low-yield nuclear weapon force (with low yields not being all that low since they include 20 kiloton nuclear weapons, the same as those dropped on Hiroshima and Nagasaki).

The most notable low-yield capabilities on the wish list include submarine-launched ballistic missiles (SLBMs) and sea-launched cruise missiles (SLCMs), which could be based on surface ships or submarines. The administration seeks to deploy low-yield nuclear weapons on both missiles to achieve the ultimate mission of the Nuclear Posture Review: to generate more flexible and tailored nuclear responses to a wide spectrum of nuclear and non-nuclear attacks against the United States and its allies. Proponents argue that incorporating more low-yield nuclear weapons into the force posture gives the United States the ability to respond to various forms of aggression with more calibrated responses on the so-called escalation ladder (and in theory, deter or defeat that aggression without escalation to the strategic nuclear level). In other words, the Trump administration hopes to generate options beyond “suicide or surrender.”

Although the aim of the low-yield SLBM and SLCM is to close this perceived “deterrence gap,” proponents of these capabilities have elided one key problem: how the adversary may perceive and react to their use. I call this the “discrimination problem.” Right now, all the SLBMs in the American inventory carry multiple — up to eight! — thermonuclear warheads. Mixing these missiles with one or several of the proposed low-yield warheads creates a very real problem: How will the adversary know which of the two is coming its way? It cannot. If the adversary sees a single SLBM headed toward it — even if that missile turns out to only be carrying a low-yield warhead — it must react as if it is facing the full brunt of American strategic nuclear use. It would be catastrophic—potentially nation-ending—to hope otherwise and be wrong.

The new low-yield, or nonstrategic, nuclear weapons envisioned in the Nuclear Posture Review would not be the first in the American inventory. There are already four types of aircraft-delivered tactical nuclear weapons in the force posture (three variants of the B-61 gravity bomb and an air launched cruise missile).

So why does the review call for additional low-yield options? In a word: Russia. The administration’s basic concern is that Russia may try to use a low-yield nuclear weapon on American or allied forces without the United States being able to successfully respond in kind. This forces America into the “suicide or surrender” dilemma of either not responding at all or escalating directly to the strategic thermonuclear level by retaliating against the adversary’s cities (or against all its nuclear forces directly). The perceived gap in American capabilities is because U.S. aircraft-delivered B61s are vulnerable to Russian air defenses, limited by the range of the aircraft on which they are deployed, and cannot deliver a retaliatory blow as swiftly as ballistic missiles can. Therefore, the Nuclear Posture Review argues, the United States needs a new capability that can penetrate Russian defenses and deliver a low-yield nuclear weapon from anywhere within minutes. The basing mode that achieves this, without requiring a host nation, is at sea. In the near term this would involve modifying existing SLBMs to carry a low-yield variant of an existing warhead (for a variety of reasons, I assume the W76), while working in the long term to deploy a nuclear SLCM.

The theory is that fielding this capability will deter Russia from its so-called “escalate to deescalate” nuclear strategy (insofar as that even exists), which is premised on the notion that using nuclear weapons early in a conflict, but in a limited way, will lead the United States to back down. If deterrence fails, low-yield nuclear options deliverable from American submarines provide a flexible and tailored response option to defeat Russian aggression.

Here’s why it would be so dangerous to deploy the low-yield SLBM in particular. America presently fields one type of ballistic missile on its 14 nuclear weapons-designated Ohio-class submarines: the Trident II D5 missile. Each Trident missile can carry up to 8 independently targetable warheads, some combination of the W76 thermonuclear warhead (100 kilotons) or the W88 thermonuclear warhead (455 kilotons). Currently, if an adversary were to detect a launch of a Trident missile from an American ballistic missile submarine, there would be no uncertainty about what is coming its way: a strategic nuclear launch of at least about a megaton of yield, perhaps 3.6 megatons. This is, without question, a strategic nuclear launch by the United States aimed at destroying the adversary’s high-value cities, or perhaps its strategic nuclear force itself (also known as a counterforce strike). By reserving the SLBM for strategic nuclear use — and only strategic nuclear use — there is no ambiguity about what a Trident launch means for both the United States and the adversary: all-out nuclear war.

But if the United States starts deploying some Tridents with a single low-yield warhead and others with eight thermonuclear warheads, all on the same submarine, how will the adversary know what is coming its way? There is literally no way to tell which warhead yield is atop the missile — no early warning system can discriminate between the low-yield warhead and the strategic nuclear warheads at launch or in flight. Early warning systems can detect the point of launch and perhaps the type of missile fired. But not even the most sophisticated system can discriminate between a W76 or W88 warhead that is set to deliver hundreds of kilotons and a warhead that looks exactly the same but is set to deliver just 20 kilotons. The only thing an adversary sees is a Trident missile launch, which could now be anywhere from 20 kilotons of damage (designed to destroy a military base, for example) all the way up to 3.6 megatons (enough to destroy multiple cities and kill millions of civilians). Even if the early warning system could see that there was only a single warhead instead of eight, how confident are we that the adversary will believe their radars instead of fearing the worst?

What does this mean? If the adversary detects even a single missile launch, it has no choice but to react as if the United States has decided to escalate to the strategic nuclear level. Even if the other side may hope or believe that the incoming warhead might just be a low-yield weapon, it must assume the worst, because the risks of guessing wrong include losing millions of people or potentially its entire nuclear force. It is unrealistic to assume and hope — in the thick fog of a nuclear war —that the adversary will wait until the warhead has landed, do a detailed yield assessment (even if 20 kilotons hits, how are they to know it wasn’t just because the second stage of a thermonuclear weapon fizzled?), and then choose not to respond because it was “only” 20 kilotons instead of 3.6 megatons.

Think about it this way: if the United States detected that Russia had launched a missile off a submarine, that carried either a low-yield nuclear weapon or 8 strategic nuclear weapons, how would it react? Would it assume it is the low-realyield option and wait for it to hit the continental United States before reacting and retaliating? Of course not. Yet this is what America is hoping its adversaries will do.

When it comes to waging a nuclear war, it is simply unrealistic to base a whole strategy on hoping that an adversary assumes the best-case scenario. The adversary’s most logical move is to respond as though full-scale nuclear war has started — which means that even if they were wrong, the end result and the consequences are the same. The use of a low-yield SLBM, supposedly built for a “small” nuclear conflict and to calibrate escalation, has now leapt to strategic nuclear war because of how the adversary is forced to react.

Furthermore, mixing low- and high-yield nuclear weapons on Trident missiles, one of the key systems the United States would use in a counterforce mission targeting an adversary’s nuclear forces, poses a particular problem if an adversary is worried about the survivability of its arsenal (even Russia may worry about this given America’s persistent emphasis on counterforce and damage limitation capabilities). Such an adversary may experience “use-them-or-lose them” pressures at the sight of a single Trident launch (doubts about their early warning system could lead them to believe many more are headed their way). An adversary which fears that the United States is about to wipe out its arsenal may have no choice but to launch everything it has before even knowing what is actually incoming. This is certainly the case if the adversary is North Korea, it might be the case for China, and could plausibly be the case for Russia.

The argument that an adversary might be nonchalant about “only one or two” Trident missiles headed its way makes delusional assumptions about how a state facing the most potent nuclear force on the planet might react. For many nations, “one or two” Trident missiles with eight to 16 strategic nuclear warheads would be life-ending. They cannot afford to be complacent or assume best-case scenarios. This is why it is so important to load only strategic nuclear warheads on SLBMs, so there is no ambiguity about American intentions and about what is being launched at the adversary. The virtue of keeping SLBMs (and intercontinental ballistic missiles) single-assigned strictly with strategic nuclear warheads is that these missiles are the signal that the United States has escalated to the strategic nuclear level.

The discrimination problem outlined here applies very specifically to mixing low-yield and strategic nuclear weapons on the same missile and same system, deployed on the same platform (in this case submarines). The same concern would apply equally to a proposal to load low-yield nuclear weapons onto intercontinental ballistic missiles.

The low-yield cruise missile may be a less bad option in this regard, since cruise missiles have different flight profiles and only carry a single nuclear warhead. An adversary is less likely to mistake a single cruise missile launch for full strategic retaliation. However, proponents still need to explain why it is necessary — other than to try to develop a bargaining chip to force Russia back into compliance with the Intermediate-Range Nuclear Forces Treaty. It’s unclear what deterrence gap the new SLCM capability will fill after the long-range standoff cruise missile with a low-yield option is developed to replace the current air-launched cruise missiles.

In trying to deter more — and lower — forms of aggression with nuclear weapons and broaden the deterrence spectrum, the Nuclear Posture Review generates real risks of spirals of nuclear escalation in a crisis or war. It tries to reintroduce the idea of a calibrated “escalation ladder” — the notion that in a conflict the United States and the adversary can have various “rungs” of very precise and controlled nuclear exchanges of varying intensities without unintentional escalation. The heroic assumptions made by the idea of such a “ladder” are too numerous to address here. But a primary one is that it erroneously assumes the United States can alone control the climbing of that ladder without the enemy getting a vote. The concept fails to consider how the very existence of ambiguous nuclear systems — is it low-yield or thermonuclear? — can blow up the ladder. While the idea of a low-yield SLBM may be attractive in a sterile game theory seminar, in a real conflict with real decision-makers, it is a recipe for uncontrollable nuclear escalation.


Vipin Narang is associate professor of political science and a member of the Security Studies Program at the Massachusetts Institute of Technology.

Image: Stratcom/Adam Hartman