Michael KreponNuclear Weapons and the Laws of War (cont.)

Note to readers: Dr. Justin Anderson continues the conversation about nuclear weapons and the laws of war with this contribution. Rejoinders are, of course, welcome. Justin is a Research Fellow at National Defense University’s Center for the Study of WMD. The views expressed in this article are those of the author and do not reflect the official policy or position of the National Defense University, the Department of Defense, or the U.S. Government. His previous contribution appears here.

The article by Gro Nystuen and Kjolv Egeland in Arms Control Today titled, “A ‘Legal Gap’? Nuclear Weapons Under International Law” begins by citing language from the “Conclusion” of the Final Document of the 2010 NPT RevCon, noting it “referred for the first time in [Nuclear Nonproliferation Treaty (NPT)] history to the ‘catastrophic humanitarian consequences of any use of nuclear weapons.’” This phrase from the Final Document is also frequently cited by states and non-governmental organizations that support the Humanitarian Impact of Nuclear Weapons initiative (HINW).

This statement is factually incorrect. “Any use” is a term that encompasses every potential use. But not every potential use of a nuclear weapon would have “catastrophic humanitarian consequences.” For example, a nuclear weapon detonated high in the atmosphere or in outer space for demonstration purposes or to create an electromagnetic pulse might have consequences for hardware – but limited humanitarian consequences. Dan Joyner’s post in ACW also notes examples such as the isolated military targets cited in Schwebel’s dissenting opinion on the ICJ’s 1996 Legality of the Threat or Use of Nuclear Weapons case.

Nuclear employment against these types of targets, distant from population centers, could destroy them without significant humanitarian consequences. Moreover, the targeting and weaponeering factors discussed within Dunlap’s “Taming Shiva” article are directly relevant to limiting the humanitarian consequences of employment against other potential military targets, such as submarine bases or other facilities associated with a state’s nuclear forces. Another example is the potential use of a nuclear weapon to deflect or break up an asteroid to prevent it from colliding with Earth, which sounds like science fiction but is a scenario currently being studied by NASA and NNSA (in part due to the meteor whose explosion in 2013 above the Russian city of Chelyabinsk injured hundreds).

The above are not arguments regarding the merits of these possible forms of employment (although I would guess even the most committed supporters of HINW would back the use of a nuclear weapon to save humanity from the same fate as the dinosaurs). They are evidence against the use of sweeping generalizations such as “any use”, whether in negotiations, diplomatic texts, or legal or strategic analyses. Supporters of this language may object to the examples offered here, stating that what they really had in mind were attacks on civilian areas or nuclear exchange scenarios where ground bursts cause significant radioactive fallout. But word choices matter in diplomacy, particularly with regard to legally binding treaties; inaccurate phrasing such as “any use” can lead to differing interpretations, imperfect implementation, and indifferent or inadequate compliance.

Both articles also state that there is a “legal gap” with regard to the possession and potential employment of nuclear weapons by the NPT’s five nuclear weapon states. I agree with both authors that current law places significant restrictions on potential nuclear employment, restrictions that are important to acknowledge and address in the design, development, training, planning, and deployment of nuclear forces. I have two objections, however, to the term “legal gap.” First, I disagree that there is a gap in existing law: the legality of nuclear weapons possession is governed by the NPT, and the legality of potential employment of nuclear weapons is governed by the law of war.

A second, related objection is that the term is one of advocacy rather than analysis. A gap is something that needs to be filled in or bridged over. Use of the term strongly implies advocacy for a new treaty putting in place a blanket prohibition against nuclear weapon possession and use. This reflects a desire for new law rather than an assessment of current law. But the suffering and mayhem on contemporary battlefields will not be resolved by efforts to ban any particular type of weapon, regardless of its effects. Ruthless regimes and violent extremists have repeatedly proven their willingness to use any means, legal or illegal, to indiscriminately wage war. Rather than attempting to negotiate a ban that would likely prove difficult to verify and impossible to enforce, efforts to reduce the threat of nuclear war should focus on developing new nuclear arms control regimes and encouraging all states to better implement current law – to include law of war limitations on the employment of all weapons, whether conventional or nuclear.

Law of war limitations to potential nuclear employment, however, do not necessarily place a specific numerical limit on the U.S. nuclear arsenal. The U.S. nuclear force is not – as Dan suggests – 7,000 hammers looking for a bunch of nails. As of September 2014 the U.S. nuclear stockpile was 4,717 weapons; over the preceding two decades the United States dismantled 10,251 warheads. If the former number may seem large to some, consider the following: The number includes both “active” and “inactive” warheads, with the latter stored in depots and not fully configured for near-term employment. In addition, the platforms for delivering these weapons – none of which are on “hair-trigger alert” – are subject, like all military hardware, to operations and maintenance cycles. Like any fleet, not every delivery system (or the weapons associated with it) is always immediately available. Given all these factors, the number of active hammers – and readily available hammers – is well below 4,717.

In addition, per the terms of New START, by 2018 the United States will have reduced its number of deployed strategic weapons to a number at or below the treaty limit of 1,550. The actual number of U.S. deployed strategic weapons will slightly differ from the figure reported under the treaty, as New START’s “treaty math” counts each strategic bomber as one weapon regardless of loadout. It is also important to note that New START does not cover or count “non-strategic” nuclear weapons. Even after taking these factors into account, however, U.S. compliance with the treaty and its continuing dismantlement of old weapons has the deployed strategic force, and total U.S. nuclear stockpile, at their lowest levels in decades. Nevertheless, in line with its NPT Article VI commitments, the United States remains open to arms control negotiations with Russia to discuss further reductions to these numbers.

Advocates of reaching low numbers in the near future may contend that the U.S. stockpile is still too large and conclude, as Dan does, that the size of the arsenal must indicate an interest in retaining nuclear weapons for a wide range of scenarios and targets both nuclear and non-nuclear. This conclusion is incorrect. Wherever possible, the United States seeks non-nuclear alternatives to meet deterrence and defense requirements. The 2010 Nuclear Posture Review Report, for example, articulates an enduring interest in assessing how new and evolving non-nuclear capabilities can lead to a “reduced reliance” on nuclear forces. U.S. policy also narrows and limits the parameters of nuclear employment planning. The unclassified summary of current presidential guidance provided by the 2013 Report on the Nuclear Employment Strategy of the United States clearly states the Commander-in-Chief “will only consider the use of nuclear weapons in extreme circumstances,” while also underlining the critical role the law of war plays within this force’s military plans:

all plans must also be consistent with the fundamental principles of the Law of Armed Conflict. Accordingly, plans will, for example, apply the principles of distinction and proportionality and seek to minimize collateral damage to civilian populations and civilian objects. The United States will not intentionally target civilian populations or civilian objects.

The integration of the law of war with military planning is thus a key factor in narrowing the potential contingencies for U.S. nuclear employment. It also places limits on targeting, prohibiting any plans that intentionally target civilians — a point also emphasized within the DoD Law of War Manual, which states “nuclear weapons must be directed against military objectives”.

These restrictions, however, should not be interpreted as endorsing or requiring steep reductions. The 2013 Report notes that implementation of current guidance, to include ensuring all plans comply with the law of war, will not lead to the United States relying on “counter-value” or “minimum deterrence” strategies – approaches often associated with limited arsenals. In contrast, the guidance emphasizes the importance of maintaining “significant counterforce capabilities” to effectively deter a range of potential nuclear-armed adversaries from attempting to coerce or attack the United States or its allies. This reflects a strategic assessment of contemporary deterrence and defense requirements supported by a legal analysis underlining the importance of focusing plans on adversary military targets – which, with regard to highly-valued assets such as nuclear capabilities, are likely to have significant active and passive defenses. In short, including the law of war as an important part of planning processes for U.S. nuclear forces informs restrictions on contingencies and targets while also contributing to the case for fielding a robust nuclear deterrent force, given the current security environment.

Those who identify and wish to fill a “legal gap” on nuclear weapons possession and use are sincere in their desire to eliminate the scourge of nuclear war. I do not agree, however, with either their assessment of the law or their conclusion that a new treaty attempting to put in place a total, global ban could eliminate nuclear risk, nuclear competition, or the threat of nuclear coercion or conflict. Moreover, legal assessments, while critical to international diplomacy, provide only partial answers to the challenges posed by weapons of mass destruction. The history of nuclear arms control agreements demonstrates the importance of bringing together legal expertise, diplomatic acumen, military strategy, and technical proficiency to develop pacts that limit or reduce nuclear forces in a stabilizing, verifiable manner. Dan, Gro, and Kjolv provide analysis and commentary on the “legal gap” debate that I disagree with, but that arms control wonks should read alongside historical analyses of force developments, detailed assessments of foreign missile programs, and other evaluations and explorations of nuclear topics that can inform their development and negotiation of future nuclear treaties.


  1. Michael Krepon (History)

    Thanks for continuing this conversation.
    I think you make a persuasive case that one can imagine a singular use, or very selective uses of nuclear weapons’ employment that could meet the criteria of the laws of war.
    Could you venture a guess as to how many detonations and targets might meet these criteria? The United States still possesses thousands of warheads. What percentage of them might be used that would be consistent with the laws of war — assuming, of course, that escalation could be controlled?
    If escalation cannot be controlled, even with singular or very selective employment, then your basic supposition would be tenuous, would it not?

    • Daryl G. Kimball (History)


      Michael Krepon asks a very pertinent question. Would be interested in your response.

      Thinking it through myself, I find the language in the 2013 Report on Nuclear Weapons Employment strategy to be less than reassuring from a practical or a legal perspective. It states that:

      “…all plans must also be consistent with the fundamental principles of the Law of Armed Conflict. Accordingly, plans will, for example, apply the principles of distinction and proportionality and seek to minimize collateral damage to civilian populations and civilian objects. The United States will not intentionally target civilian populations or civilian objects.”

      It states that “all” plans, not just some must be consistent the the Law of Armed Conflict. It states that the U.S. will “seek” to minimize collateral damage. When it comes to nuclear weapons use, collateral damage is fairly certain if there are more than several weapons involved (let alone the risk of escalation). Some variations on the war plan involve hundreds of warheads.

      It states that the United States will not “intentionally” target civilian populations or objects. The Russian MoD building is ostensibly a military target, yet it lies in the heart of a city. Russia’s western ICBM fields are military targets, yet the fallout from a U.S. counter-strike would adversely affect civilian objects, namely civilians.

    • Justin A. (History)

      Michael, thanks for posting this and for these questions. My views here (and in replies below) are my own and not those of NDU or the USG.

      I believe that all weapons deployed by the United States go through a planning process that adheres to law of war principles, per the 2013 Report on the Nuclear Employment Strategy of the United States — plans where potential employment would occur due to “extraordinary circumstances” where the United States or an ally was facing a (very) dire threat. These circumstances would likely be linked to a nuclear or biological weapons threat (and very likely in imminent form), per the 2010 Nuclear Posture Review Report.
      To halt a mass casualty/mass destruction attack, the law of war permits consideration of significant amounts of force (while at the same time would likely counsel that not every available weapon or type of attack would be acceptable). Let me offer the following: If a significant number of targets associated with these types of WMD exist, and if conventional methods and means for their timely & assured destruction do not, then there is no inherent law of war rationale that would rule against possessing a robust and diverse nuclear arsenal (for the NPT five). Importantly, of the ~4700 weapons in the U.S. arsenal as of late 2014 (per numbers published by State Dept), a smaller number than that are actually in the field at any given time, and a smaller number than that stand ready for very rapid employment. However, while U.S. planning must prepare for a range of possible contingencies (again, all within the realm of “extraordinary circumstances”) I would add that I believe the bar for employment of a single nuclear weapon, given law of war principles such as discrimination, is very high; the bar for use of higher-yield weapons is higher still; and the bar for employment of a number of weapons (rather than a single weapon) is higher than that.

      With regard to escalation, there are a number of factors beyond the law of war that should be considered, of which I think the most important is an assessment of the cost/benefit calculus of an adversary (and how this relates to their possible use/employment of nuclear force). However, the law of war principle of proportionality does ask decision-makers and commanders to consider the consequences of actions beyond an immediate action/attack – and to calibrate their use of force accordingly, in no small part so that a later, lasting peace is possible. I believe discussions of this principle usually take a longer time scale into account than is normally thought of in terms of nuclear escalation, but the idea of factoring in how an adversary might react to your action — and thus calibrating your own action in order to reduce the possibility of bad outcomes — is there. But I think I would suggest that considerations of escalation require multiple inputs to make an informed decision, of which legal counsel is important but does not stand alone. And a separate but related observation is that just because a course of action is legal does not necessarily make it wise. – Justin

  2. Jonah Speaks (History)

    “a nuclear weapon detonated high in the atmosphere or in outer space for demonstration purposes or to create an electromagnetic pulse might have consequences for hardware – but limited humanitarian consequences.” The humanitarian consequences of EMP could be quite serious – shutting down electricity over wide areas for a long time, possibly leading to mass starvation. You should pick a more benign example.

    Even with a better example, would a “benign” first-use nuclear demonstration really be so benign? Suppose I hold a machine gun and speak angry words to a crowd of people. As proof of my seriousness, I fire a single bullet over everyone’s heads, killing no one. Is my demonstration benign, or is it a threat? If someone in the crowd also has a machine gun, what is the likely consequence?

    Merely to speak of a demonstration implies first use. If the other side uses nuclear weapons first, there is no need to first provide a demonstration, one merely proceeds to a second use. Would you be willing to ban the first use of nuclear weapons? Would you be willing to accept restrictions on what would constitute a proper second use of nuclear weapons?

    • Julius Kozma (History)

      Regarding the detonation of NW high in the atmosphere or in outer space for demonstration purposes or to create an electromagnetic pulse – the Partial Test Ban Treaty signed by the US prohibits carrying out any nuclear weapon explosion, or any other nuclear explosion…in the atmosphere; beyond its limits, including outer space. Also the United Nations General Assembly adopted resolution 2222 (Outer Space Treaty) prohibits States Parties from placing in orbit around the earth any objects carrying nuclear weapons or other weapons of mass destruction. In addition to prohibiting nuclear testing in space, the Outer Space Treaty also prohibits Parties from engaging in military maneuvers on celestial bodies, conducting nuclear tests on celestial bodies, installing weapons systems or constructing military bases on celestial bodies.
      I can see a possibility that the States agree on an exemption from the Outer Space Treaty in case of an eminent danger to humanity, but demonstration purposes are prohibited.

  3. Kjølv Egeland (History)

    Thank you for commenting on our article in ACT, Justin.
    On the legal gap: It is not particularly important to me what we call the ‘legal gap’. (When asked this exact question by the Polish delegation to the OEWG in Geneva on Monday, Stuart Casey-Maslen, invited as an expert by the Chair, said he could settle for ‘a gap in the law’.) The point is that, for many, it seems illogical that five states, and five states only, are allowed to possess – and maybe even use – weapons of mass destruction that are more difficult to use in compliance with the rules of IHL than weapons that have already been banned because they are difficult to use in accordance with those very same rules. On the idea of a ban: I don’t think anyone supporting a ban treaty believes it would ‘eliminate nuclear risk, nuclear competition, or the threat of nuclear coercion or conflict.’ Rather, it would be one of many measures, including the arms control measures you mention, that would help bring about a world without nuclear weapons.

    • Justin Anderson (History)

      Kjolv, thanks for the questions. I enjoyed your article; while I disagree with your conclusion, I think it is a good analysis. In regard to the “legal gap/gap in the law” (and I don’t agree with latter term either), two things: 1) I think biological weapons are less discriminate than nuclear weapons; 2) The NPT is a package deal, with differing obligations for nuclear and non-nuclear parties. I recognize that some object to the terms and conditions, and some object to the pace of implementation of Art VI (I disagree on both counts). But I do not see a lack of logic in the deal.

      In regard to the relationship between a ban and arms control, my own view is that they are incompatible. The latter accepts that some number of weapons can play a role in national security and strategic stability. In general I think bans are too hard to definitively verify and are too fragile to bring about lasting stability.

  4. Michael Krepon (History)

    The US SIOP for 1956 has been declassified, and Alex Wellerstein, as is his want, has clarified the fallout consequences of all of these detonations. His visualization and depiction are helpful. Qualifiers are in order: The US stockpile looks much different than in 1956. Yields have been scaled down significantly. While escalation control was even more problematic in 1956 than it is today, the detonation of every single weapon in the SIOP is questionable. And some detonations would be above ground as well as at surface level. All quibbles. Have a look here: