Michael KreponApplying Jus in Bello to the Nuclear Deterrent

Dear readers: Here is another in a series of posts on the laws of war and the use of nuclear weapons. This contribution is from Dr. Justin Anderson, a Research Fellow at National Defense University’s Center for the Study of WMD. Justin previously worked at SAIC in support of the Air Force, STRATCOM, OSD General Counsel/International Affairs, and the OSD Office of Treaty Compliance. From 2009-2010 he was Editor of the DoD Law of War Manual. 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.

On December 7, 2015, the UN General Assembly passed A/RES/70/50, titled “Ethical imperatives for a nuclear-weapon-free-world,” by a vote of 132-36. Co-sponsored by Austria and several other states central to the “Humanitarian Impact of Nuclear Weapons” movement, the resolution charges that any use of nuclear weapons would inherently violate the laws of war, stating: “it is inconceivable that any use of nuclear weapons, irrespective of the cause, would be compatible with the requirements of international humanitarian law.” The resolution’s claim that there is an airtight legal case against any form of nuclear employment merits close assessment by jurists, strategists, and policy wonks, particularly in states fielding nuclear forces.

The law of war consists of two parts: jus ad bellum (justice of war) and jus in bello (justice in war, i.e., in the conduct of war). A great deal of literature on nuclear weapons focuses on a leader’s decision to “push the button” (or not); without explicitly mentioning the law of war, these writings often implicitly address jus ad bellum issues. While important, this focus on the highest levels of decision-making leaves a gap with regard to how states apply key jus in bello principles – such as distinction, military necessity, humanity, and proportionality – to their development and deployment of nuclear-capable forces (and the personnel who operate them). This post will briefly describe these principles and provide a “wavetops” assessment of how they apply to the U.S. nuclear deterrent – a force that directly reflects the full implementation of arms-control treaties and an enduring commitment to the law of war.

The principle of distinction expressly divides all persons and objects between those that are legitimate targets with regard to the use of armed force and those that are not, and prohibits attacks directed against the latter. The principles of military necessity and humanity are interlinked. Military necessity permits the use of force against legitimate targets in pursuit of military objectives while humanity dictates that this use not be deliberately excessive and/or pointlessly cruel (e.g., by adding poison to weapons expressly to prolong suffering or defeat medical treatment). A closely related principle is proportionality. This is not, as it is sometimes misunderstood, a principle dictating that a response to an attack be in-kind – that is, “proportional” to the amount of force used against you. Instead it refers to an assessment of whether the force contemplated for an attack, operation, or campaign is proportional to its intended military ends. In time of war, for example, every adversary military base, depot, and vehicle is a legitimate target – but it is unlikely the total destruction of all of them is necessary to secure victory. The principle of proportionality asks the commander to calibrate the use of force to a level that successfully achieves military objectives, while not exceeding a threshold that results in wanton destruction or deeply harms the prospects for a future peace. Proportionality is context- and situation-dependent; if the military requirements are great, it does not rule out significant or even massive use of force.

State adherence to these principles requires a significant investment of political and economic capital to equip militaries with the means and methods to identify and strike legitimate targets while also limiting the risk of harm to civilians and civilian objects.

These are core principles of the law of war; how does the United States apply them to its nuclear deterrent?

The U.S. government publicly states that all nuclear employment plans “must [be] consistent with the fundamental principles of the Law of Armed Conflict.” U.S. nuclear forces adhere to the principle of distinction in terms of targeting policy (as a matter of law and policy, the United States does not target civilians or civilian objects); target identification (U.S. investments in intelligence, surveillance, and reconnaissance technology are critical to correctly identifying legitimate targets); and guidance technology (hardware and software ensures delivery systems are highly accurate; U.S. ballistic missiles, for example, have a very low circular error probability).

U.S. employment doctrine and planning also integrates the principles of military necessity, humanity, and proportionality. Should the United States or an ally, for example, face an imminent nuclear attack, the U.S. military might advise the president that preventing the attack would require a rapid strike, launched at a distance, using munitions that would completely disable or destroy – rather than merely degrade – the belligerent forces preparing the attack. These requirements might rule out available conventional options; in this scenario, a U.S. nuclear strike would be a legitimate response due to the military necessity of completely neutralizing the target in order to prevent a catastrophic, mass-casualty attack against the United States or an ally.

Importantly, however, while the law of war would justify this use of force, it would also require that it be carefully calibrated; the above scenario would not be a license to unleash Armageddon. In an Air Force Law Review article titled “Taming Shiva: Applying International Law to Nuclear Operations,” Air Force Major General (ret.) Charles Dunlap – who at the time of writing was the principal legal advisor to the commander of U.S. Strategic Command – discussed a number of ways to adjust the effects of a nuclear strike, to include: “reducing weapon yield, improving accuracy through weapon system selection, employing multiple small weapons (as opposed to a single, large device), [and] adjusting the height of burst – [through utilizing some or all of these options] collateral damage can be minimized consistent with military objectives.” Many discussions of potential nuclear weapons employment appear to assume all attacks have the same effects; the measures described by Dunlap, however, can exponentially reduce the potential risk to civilians and civilian objects associated with any theoretical strike.

The U.S. commitment to integrating the law of war across all military activities, to include nuclear operations, willfully limits potential U.S. nuclear employment to “extreme circumstances” and also conditions what form – in terms of the target, delivery system, and munitions selected – this employment might take.

This is a brief assessment of a critically important topic, but one I hope helps catalyze discussion of the law of war and its relationship to the development and deployment of nuclear delivery systems and weapons in all states fielding these forces.


  1. Reid Byers (History)

    Ius in bello issues become ius ad bellum issues when the subject is nuclear war. Even considering the use of the weapons raises all the ad bellum questions. The genie miraculously got stuffed back in the bottle after 1945, but allowing the genie out, under any circumstances, means that every country will feel more free to make use of nuclear weapons. So working on how to use them in a proportional manner is certainly an activity with dangerous ius ad bellum implications. No?

    • Jonah Speaks (History)

      To amplify this point, the important question is not, how many people will this first nuclear use kill, and will they be military folks or civilians? Rather, the important issue is, how does this war end? It would be utterly foolish to suppose that the result would be “de-escalation” or some other happy ending, and good reason to suppose that the unhappy ending could be akin to a full-scale nuclear war.

      Nevertheless, IF a nuclear use is contemplated, it is better to start with something small and proportionate, and hope that the enemy responds with something equally small and proportionate, rather than something worse. Despite this limited first move, there is no serious basis for supposing with anything like near certainty that the resulting nuclear war will stay “limited.” To be sure, there is also the nuclear precedent for future wars, assuming we all survive to fight again.

  2. Steven Hayden (History)

    The DPRK has not received civilized treatment from UN or USA. The UN and USA bombed every building in DPRK in Korean War which includes all civilian buildings as well.About a third of civilians were killed from UN bombing. Now talk to them about the rules of war.Then sanctions for last 65 years help inflict pain suffering upon DPRK .It difficult to demand mercy from those who have not received it . Tactically nukes compensate for DPRK deficits in conventional warfare ability.

  3. Jonah Speaks (History)

    “Should the United States or an ally, for example, face an imminent nuclear attack, the U.S. military might advise the president that preventing the attack would require a rapid … U.S. nuclear strike.” With all due respect, this scenario sounds like sheer fantasy. Unless you know the nuclear order has already been given, or you see the nuclear missiles coming your way, how do you know any of this with enough certainty to contemplate a nuclear first strike? If you actually know the enemy has irrevocably committed himself to a first strike, your response is technically a second strike. If you don’t know the enemy’s nuclear plan, you are advocating starting a nuclear war first, for fear that your enemy may start a nuclear war first.

  4. Suraya Omar (History)

    It appears that what makes nuclear weapons unique – and powerful in comparison to conventional weapons – is their capacity to be fantastically destructive, such that a single bomb threatens an entire city without the aggressor committing to a Dresden-like operation. Further, minimizing the risk to civilians will almost always have major limitations based on the subsequent release of radiation. I’m having trouble imagining a situation in which the U.S. (for e.g.) would execute a “carefully calibrated” attack on a specific target with nuclear weapons. Under what circumstance would cutting back on needed ordnance or bomber sorties justify the paradigm shift inherent to reintroducing the use of nukes? Or is there an argument that the effect of nuclear weapons can be tuned and mitigated enough that they don’t have to be game changing?

    • Michael Krepon (History)

      I think it’s fair to say that the number of targets that would appear to meet these criteria are far less than the two largest stockpiles.
      Even targets that meet these criteria require additional provisos.
      Take, for example, ICBM silos or garrisons that are not downwind of large metropolitan areas. Targeting a single silo could meet the criteria. But it’s unlikely that just one silo would be struck. Far more likely is that many or all silos would be struck. In this event, escalation control gives way to damage limitation.
      The cumulative effects of nuclear warfare at this level of magnitude make a mockery of just use.

    • Cameron (History)

      At the point where your intention is to destroy a city to destroy a target in it, you’ve reached a jus in bello issue distinct from the type of weapon which you would be using to carry out the mission. There World War Two justifications for strategic bombing – civilian morale and infrastructure destruction both failed, and the resulting destruction was the second most abhorrent thing about that war. Throwing around -wide spread- destruction on that level, unless at a purely military target (B-52 raids in the first Gulf War against military positions, ie.) has become taboo.

      The use of a nuclear weapon in a similar circumstance would require both a desperate need for the guaranteed destruction of a military target, and a threat on par with the use of a nuclear weapon. The only reasonable targets that I can think of are ICBM silos, SLBM tender facilities, or biological weapon staging, that are about to be or have begun to be used. In which case you’re in a nuclear war, and you’ve moved from nuclear use in conventional fighting to a failure of deterrence. Anything else, from the US perspective (and ideally our nuclear umbrella allies), conventional strike can handle

    • Jonah Speaks (History)

      In the linked article “Taming Shiva,” Dunlap (1997) notes: “while the U.S. does not target populations per se, it reserves the right to do so under the limited circumstance of belligerent reprisal. The U.S. (along with other declared nuclear powers) insists that Protocol I to the Geneva Conventions does not apply to nuclear weapons. Hence, prohibitions contained in Protocol I forbidding reprisals against civilians are not, in the U.S. view, applicable to nuclear operations.” (p. 163) Is Dunlaps’ statement still true?

      Would such reprisals include bombing enemy cities in retaliation for bombing U.S. (or allied) cities?

  5. Allen Thomson (History)

    Silo attacks would very likely use contact fuzing, i.e. ground bursts, and produce large amounts of fallout. In the Cold War context, the results of that would have been extreme. See the FEMA study NAPB-90 and remember that the mean lethal dose is about 600 roentgens: