Jon WolfsthalThe 1st Nuclear Ban Draft is Out

Long time reader, first time Wonker.  My thanks to the Wonk-Supreme for allowing me to contribute.

The president of the negotiations for a nuclear weapons ban convention has just released her first draft of the proposed treaty. As will become readily apparent, I am not an expert on international law, but I am a keenly-interested observer of the effort to negotiate a ban on nuclear weapons. I am also an admitted ban skeptic, not because I want to keep nuclear weapons but because I have am concerned the ban diverts scarce resources and attention from readily-identified blockages to nuclear reduction and elimination. However, as a life-long supporter of nuclear reductions and disarmament, I support efforts to harness global frustration with the pace of nuclear elimination to advance disarmament in a way that advances global security. So if there is to be a ban, I want it to be additive.  An initial read of the just released draft text raises new questions and concerns, some more important and less easily resolved than others. Keeping in mind a first draft is just that, a first draft and will undoubtedly be improved and clarified as talks continue, my concerns and questions fall into the following categories:

1) Stationing – NATO countries agree to share the burden of nuclear deterrence in a variety of ways. One is that some NATO countries agree to station forward deployed nuclear weapons on their territory to enhance the credibility of reassurance and deterrence. The draft treaty would prohibit this by banning any member state from allowing nuclear weapons to be stationed on their territory. I have long supported the withdrawal of forward deployed nuclear weapons form NATO as part of a verified agreement with Russia to return its own tactical nuclear weapons to central storage. However, the ban treaty would put NATO states at an immediate disadvantage and without Russia having to modify its deployments or verify their storage away from the battlefield. Already skeptical of the realism of the ban convention, this clause will add to the concern the United States and NATO countries have expressed toward the ban. If the Treaty enters into force, could a country seek a new legal ruling from the International Court of Justice (ICJ) that NATO states with nuclear weapons on their territory are violating international law?  Would they have to choose between compliance with a ruling and NATO membership?

2) Testing – Article I of the draft treaty duplicates the language of the Comprehensive Nuclear Test Ban: “Each state undertakes never under any circumstances to . . . [c]arry out any nuclear weapon test explosion or any other nuclear explosion.” While the authors almost certainly intended to be supportive of efforts to ban testing, the terms of the CTBT are embedded in a specific framework that includes provisions for monitoring and on-site inspection, establishment of the Comprehensive Test Ban Organization, the International Monitoring System, and an understanding among the nuclear weapons states regarding permitted activities such as subcritical nuclear testing. Replicating the language here creates an uncertain relationship between the two obligations, raising the possibility that the ban treat will do harm to CTBT.

3) The Nonproliferation Treaty (NPT) and the Ban – Article 19 of the draft treaty states that the “[c]onvention does not affect the rights and obligations of the States Parties under the Treaty on the Non-Proliferation of Nuclear Weapons.” The NPT recognizes that states who conducted a nuclear test before 1967 will continue to possess nuclear weapons and “[e]ach of the Parties to the Treaty undertakes to pursue negotiations in good faith on effective measures relating to cessation of the nuclear arms race at an early date and to nuclear disarmament, and on a treaty on general and complete disarmament under strict and effective international control.” If the ban convention enters into force and takes on the weight of a new international legal norm that would apply even to states who have not joined the agreement, then would the NPT’s allowance of nuclear possession stand or be reversed? Can the ICJ find nuclear possession by NPT nuclear weapon states illegal? This is not a criticism of the draft text, but a serious question that would need extensive analysis by legal scholars and would need to be reflected in the negotiations. It is also a central concern of most, if not all nuclear weapon states and their treaty allies with regard to the ban and its impact.

4) The Role of the International Atomic Energy Agency (IAEA) – The draft treaty would erase almost 20 years of progress on IAEA safeguards.  We have come a long way since the INFCIRC 153 was the standard upon which the IAEA was based.  The IAEA safeguards system has adapted to modern realities and technology through the creation and broad implementation of the Additional Protocol (AP). Some non-nuclear weapon states have resisted adopting the AP as a new standard for safeguards not wanting to take on new obligations unless the nuclear weapon states do more to fulfill their commitments to general and complete disarmament. Putting aside both parts of this debate, adopting a standard for verification that is less than the best the IAEA and member states have achieved undermines nonproliferation and the goal of disarmament. Some nuclear weapons states will see this choice in the ban convention as another sign that it is mainly a mechanism to push obligations on nuclear weapon states and their allies.

5) Entry into Force – The draft treaty as drafted would enter into force once 40 states join. It sets no sub-category of states regionally or technically related to their status as alliance members or states capable of producing nuclear weapons. Thus, 40 countries relatively unaffected by global security or nuclear security considerations could be in a position to set a new normative legal standard that would apply to all states. The thinking behind this, as opposed to other possible standards will be a question worthy of attention when the next round of talks begin this June.

I can see ways a ban would be additive and helpful to U.S. and global security and advance disarmament. However, a ban treaty that would undermine the affirmative and voluntary basis for reassurance among allies and undermine the standards set in the CTBT and for IAEA safeguards is one that should be viewed with great concern. Maybe my parents were right and I should have become a lawyer, but barring a time machine I will be asking my legal scholar friends for their opinions on what these questions mean for the future of the ban, the NPT and disarmament if implemented as drafted.


  1. Paul Ingram (History)

    Jon, don’t your concerns only apply if NATO states are bound by the treaty? And that would be their choice and would not lead to proliferation pressures. The only states bound by the treaty would be signatories. And at this stage a bigger that to the NPT may be a failed or drawn out Ban Treaty process that leaves negotiating states even more frustrated.

    • Jon Wolfsthal (History)

      This gets to one of the biggest central issues of the proposed ban – will it take on the weight of international customary law and become binding on all, even non-signatories or will it only apply to those who voluntarily take on its obligations? At least some ban proponents want, and some NWS fear that the Treaty will be configured in such a way as to make the obligations binding regardless once it gain broader adherence. I have humbly suggested that one way to avoid this and the biggest dispute between ban advocates and opponents would be to include in the treaty an explicit article that it is not meant nor should it be used as a basis to convey a binding obligation on anyone except willing participants. But again, I am eager for someone with international legal treaty experience to weigh in here.

    • Marcus (History)

      Hi John

      Like you, I’m a long time reader but first time writer. Full disclosure, I’m not a lawyer but I know a spot about treaty law.

      A critical point about whether it will become customary international law is what is known as “specially affected states”. It’s a central consideration in whether or not something could be considered customary international law. Basically, unless those states that are particularly affected by a treaty become bound by the particular restriction that one is seeking to prove as being customary, it’s then not generally considered to be customary.

      Quick explanatory example, consider the treaty in question. Those states that would be most affected by the treaty are probably states with nuclear weapons. These countries are most likely the the set of specially affected states in a consideration of whether a nuclear ban is customary as a nuclear ban treaty has little legal affect for countries that do not have weapons and are already party to the NPT. It would require at least some (a good selection of them too, NPT and non-NPT) if not all of them to sign up for it to likely pass this test if the question of whether the ban would become customary law.

      Another example (that if memory serves, comes up in the ICJ decision on nuclear weapons) is sea law. The example given is that it would be hard for a maritime law to be considered customary if the UK we’re not a party to it it. The UK has had and to some extent continues to have such a significant role in maritime “stuff” that something would be hard to be considered customary if they were not a party of adherent to a particular rule or convention.

      If you go to the ICJ ruling on the use of nuclear weapons there’s a reasonably clear explaination of the principle. And while a future ruling will really depend on the composition of the bench, I’d say the same arguments there in the 1996 would be repeated in any case on a nuclear ban treaty.

      So unless something crazy happens, I’d say that a ban being declared customary law is very unlikely for a very long time.

      Hope that’s clear. Thanks again for the post.

    • William Alberque (@walberque) (History)

      It’s important to be aware of the concept of Persistent Objector and its interaction with customary international law. Paul Deen from State and Stephen Hill from NATO are good resources should you want to know more.

  2. J_kies (History)

    Jon; I have a ongoing question; how would the treaty materially impact current rogues from or scofflaws of the international norms. Specifically; the DPRK and Israel seem to be significant outliers.

    States post NPT such as Pakistan and India do not show signs of relenting from their destructive behaviors towards norms either.

    How does a treaty propose to show these nations that adherence to norms is in their survival interests?

  3. Michael Krepon (History)

    Welcome aboard! Your gravatar makes you look even younger than mine makes me look younger. Have no issue with that.
    My initial thoughts after reading the first draft are that the language is carefully and professionally drawn; your suggestion about the Additional Protocol standard is well taken; and that, as drafted, the Ban Treaty’s language reinforces the NPT more than it impells nuclear disarmament. As such, criticisms from the arms control community may be overdrawn. But then again, I’m not a lawyer either.

    • Jon Wolfsthal (History)

      Michael: I do respect the depth and professionalism of discussion at the first session, and of the President of the Conference in her initial draft. However, the bulk of the objections IMHO are not centered on whether the text replicates or distances itself from the NPT. Yes, there is a concern about regime and treaty shopping, but I think the larger concern is two fold – One, that the work going into the ban convention would be better spent raising concerns with Russia over arms control cheating, India and Pakistan over arms racing, China over lack of transparency and putting great pressure on North Korea. Second, there is legitimate concern that ban proponents are trying to achieve through larger international legal mechanisms a basis to compel states to abandon their reliance on nuclear deterrence and reassurance without their full consent. It is the last poin on which there is great legal debate and need for broader (and self-) education.

    • William Alberque (@walberque) (History)

      What do you think the impact of the draft Treaty will be on the NWFZ Protocols and the status of Security Assurances within the UN and NPT? Does this make them illegal? Can CIL make UN matters illegal (for lack of a more sophisticated way of putting it)?

  4. Jonah Speaks (History)

    My review suggests that any nuclear-armed state ratifying the treaty has 90 days plus 18 months to get rid of all its nuclear weapons. (See articles 1 and 2 and Annex.) Did I get this right?

    Aside from the shortness of time to dismantle the weapons, there is a more serious problem here: Unilateral nuclear disarmament. The disarmament obligation is unconditional, regardless of how many or which other nations may still possess nuclear arms. Unless this obligation is conditioned on some kind of mutual and verifiable constraint on the nuclear arms of all nations, this is a complete nonstarter.

  5. Susan Burk (History)

    Jon et al: This is a good start to what promises to be a fascinating discussion. I was wondering if anyone knows the significance of the December 5, 2001 date in the draft.

    • Jeffrey Lewis (History)

      Belarus, Kazakhstan and Ukraine agreed to the elimination of all nuclear strategic offensive arms located in their territories seven years after the Lisbon Protocol took effect on December 5, 1994.

      Sarah Bidgood realized it before the rest of us. She was the office hero of the day.

    • Pavel (History)

      It’s, in fact, an odd date for a nuclear ban treaty – START obligations say nothing about nuclear warheads. They could have stayed in Belarus, Kazakhstan, and Ukraine after all launchers were eliminated. In reality, though, everything was moved to Russia by November 1996. See the discussion at

  6. Keve (History)

    Every nation has right to defend themselves and this NPT is grossly insufficient to protect non nuclear states from aggression from nuclear haves. No nation is going to take NPT seriously except to buy time, especially when nuclear haves been on war tracks more than any of non nuclear states. Unless nuclear haves show the benefits of eliminating the nuclear inventory and elimination….this system of nuclear free world will not work.

    Non nuclear states majority must create incentives and enforcement to reduce and eliminate nukes from nuclear weapons states. Incentives could be lower tax credits for economic trading for non nuclear weapons states and enforcement would be international annual economic and peace meetings without the invitations of nuclear weapons states, living them out of largest consumer global market decision making process of non nuclear weapons states. Nuclear free world is not the goal of nuclear haves, and point was very clear when nuclear haves blamed the other nuclear haves; there should be price nuclear haves must bear for endangering the world.

  7. John Carlson (History)

    Compliments to Jon for being so quick off the mark. He is absolutely right, having one treaty duplicate another is a recipe for confusion, if not damage. For this reason it’s a bad practice. Jon has pointed to two examples:
    (1) The prohibition on testing – without any cross-reference to the CTBT (apart from the preamble) this creates questions about the relationship with CTBT provisions, e.g. regarding the International Monitoring System. We could be in the anomalous situation that a state whose ratification is required for CTBT EIF could join the ban treaty and say its obligations are satisfied, without ratifying the CTBT, leaving it still in provisional operation.
    (2) IAEA safeguards – the draft sets a lower safeguards standard than the NPT. The NPT does not refer to INFCIRC/153, but requires safeguards in accordance with the IAEA’s Statute and “the Agency’s safeguards system”. This formulation allows for the safeguards system to develop over time, the Additional Protocol being a clear example of this. However the draft (Annex para 2) freezes safeguards at INFCIRC/153. Since Art.19 of the draft preserves the rights and obligations of NPT parties, presumably the Annex applies only to states that never join the NPT. However it is unclear with respect to nuclear-weapon states – if they join the ban treaty, what safeguards standard would apply? According to the ban treaty, it would be a lower standard than other NPT parties, which doesn’t make sense. In practice I expect future disarmament treaties will have their own (rigorous) verification provisions.

    Regarding Jon’s question, could NWS join the ban treaty but retain their right to nuclear weapons under the NPT, the answer is “not really but …”. If a NWS disarms in accordance with the ban treaty, it would then have to renounce the ban treaty before it could attempt to rebuild its nuclear weapon program. The question highlights a shortcoming in the NPT – there is no mechanism for the NWS to be re-classified as NNWS. The situation is clear for a non-NPT party, it can only join the NPT as a NNWS. But for a defined NWS there is no mechanism for it to be redefined. The NWS would have to unilaterally declare itself as a NNWS and accept all the obligations of a NNWS. I guess it’s going to be a while before this arises as a practical issue.

    Finally, on Jon’s question, could the ban treaty establish customary international law? Customary international law is evidenced through the general and non-contentious practice of states. In circumstances where a major issue of substance is disagreed – such as the development and possession of nuclear weapons – there is no way 40 treaty ratifications could be taken as evidence of customary law. Even if there were say 180 ratifications the fact that the states possessing nuclear weapons all stayed outside the ban treaty would make it difficult to argue that the treaty represented general practice – the practice of the weapons-possessing states could not be ignored. What is a more compelling argument however relates to a specific aspect of the ban – whether customary international law prohibits the use of nuclear weapons. Since the only use occurred in 1945, the general practice of states over the subsequent 7 decades could suggest these weapons are not to be used. The ICJ considered this question in its 1996 Advisory Opinion and could not come to an absolute answer, but said the use of nuclear weapons would generally be contrary to the rules of international law and must comply with international humanitarian law, something that would be extremely difficult. So, a widely accepted ban treaty would be additional evidence of customary international law against the use of nuclear weapons.

  8. Paul Ingram (History)

    The bigger question for those of us in nuclear weapon states, in my opinion, is not whether this Ban process is a good or a bad thing for strategic stability, other international regimes, or world peace. But rather, now that it is under way and looks set to be agreed by many states, enough probably to achieve entry into force, how can we build models of cooperation that minimise the damage and maximise what opportunities there may be? In other words, lets be pragmatic. This Ban process has already got a lot further than many expected, largely because of the level of frustration with a lack of progress and the awareness that the importance of progress appears to have been neglected by so many of the most important states to the process. We’re not in the game of doing things as effectively as we can, and this process may not help. But IMHO that’s not now the point. We all have to ask ourselves how we can best make the most of the reality of the situation.

  9. Heather Williams (History)

    The draft seems to be heavily weighted to the preferences of civil society and a small number of states, particularly the low number for EIF. This would seem like the objective is to get an in-force treaty ASAP, so as to have a “finished product” to point to and claim victory. However, such a treaty would have minimal impact on nuclear policies not only because of the small membership but also because it continues to ignore the security context and states’ motives for nuclear deterrence policies. The one benefit of having a shotgun-treaty is largely symbolic and to make the case for a normative and IHL change towards nuclear weapons, see page one paragraph 4 of the draft (cough: challenge the 1996 Advisory Opinion). All the same, the underlying reasons behind HINW and the ban treaty remain- frustration with lack of progress towards disarmament- and will have to be addressed at some point regardless of what happens with the ban.

    And now off to debate Jon about LRSO and GBSD. Jon, was writing this your break from debate prep?

  10. Daryl G Kimball (@DarylGKimball) (History)

    As with every draft treaty text, there are issues that have to be ironed out and language that has to be reworked. Though this text seems to be a good reflection of the views expressed at the first negotiating session, the initial draft text seems to be in need of some clarifications.

    Jon raises some good questions, among them the ban on nuclear testing and the CTBT. As Jon and John Carlson note, and as I tweeted yesterday, the language in the new draft’s prohibition on nuclear testing draws from Art. I of the CTBT. But apart from the preamble, the new treaty draft does not explain the relationship with CTBT, specifically states obligations (and rights) regarding CTBT verification and the monitoring system.

    While the new treaty draft text does have a line in Art. 19 on its relationship to the NPT, Article 19 does not make reference the CTBT. That would seem to be an omission that can and should be fixed.

    We had circulated a working paper at the end of the first negotiating session on the prohibition treaty in March suggesting the new treaty does not need to include a ban on testing given that the CTBT, which has been signed by 183 states, does so. See:

    But if the new prohibition treaty is to include a prohibition on testing, I would think Art. 19 should include language making it clear that the new treaty does not limit or detract from states’ obligations under the CTBT and that the adherence to the CTBT is essential to ensuring compliance with any treaty prohibiting nuclear weapons test explosions. Though the CTBT has not yet entered into force, it has established a norm against nuclear testing and the international community has already recognized that signature on the CTBT establishes a legal obligation not to conduct a nuclear test explosion.

    On safeguards, I too am disappointed that the prohibition treaty does not properly recognize the importance and value of the improved safeguards standards, particularly through the Additional Protocol, as they related to NNWS. After all, the JCPOA makes it abundantly clear why the additional protocol and special inspections are so important. There are 172 NNWS that already have comprehensive safeguards with the IAEA; of that number 124 have in place additional protocols. Most of the negotiating states are also members of a nuclear weapons free zone, which obligates each of them to have a safeguards agreement with the IAEA. In other words, why not raise the standard a bit?

    However, as I read Article 5 of the new treaty text, it does allow for states parties to adopt and annex new verification measures (yet to be developed) that would be necessary to verify the irreversible elimination of nuclear weapons by the current set of nuclear armed states. This suggests to me that the drafters envision a case-by-case approach to verifying future disarmament by the existing nuclear armed states — which would seem to me to be a pragmatic and flexible approach.

    On a different aspect of the CPNW draft text, I am still thinking through some of the practical and political implications of the provisions in Article 9 on “Meetings of States Parties,” which could have some interesting effects on the dynamics of the international discussions and debate on forward looking disarmament issues.

  11. John Carlson (History)

    My previous comment dealt with one safeguards issue with the draft convention, the limiting of safeguards to those required by INFCIRC/153. However, there is a more serious drafting problem, with Article 4.

    There is an inconsistency between Articles 3 and 4. Article 3 requires a party to accept safeguards as provided for in the Annex (discussed previously). However, Article 4, that applies to a state that has nuclear weapons but disarms prior to joining the ban convention, is more restrictive than a normal safeguards agreement, allowing the IAEA access only to locations or facilities “associated with a nuclear weapon program”, and access elsewhere only on a case-by-case basis. It is not clear what the drafters intend here. The IAEA should have access anywhere in the state as required to verify the correctness and completeness of the state’s declaration of nuclear material, subject only to the state’s right to ask for managed access to protect confidential information (INFCIRC/153 paras 5, 76 and 77, AP Art. 7). Maybe the drafters intend that Article 4 will apply until the IAEA has verified the state’s initial inventory, and normal safeguards will apply after that. But the IAEA can’t verify the completeness of the declaration without also undertaking verification at locations and facilities not declared to be associated with the nuclear weapon program – the Agency needs access to all nuclear material in the state in accordance with usual safeguards practice.

    Article 4 as drafted is unsatisfactory. The inconsistency between Articles 3 and Art. 4 should be rectified. If Article 4 is meant to give the IAEA authority to do everything it needs to verify the completeness of the initial inventory, including access to sensitive locations (if necessary under managed access), it should say so. As currently drafted Article 4 compromises the Agency’s safeguards authority.

  12. David Koplow (History)

    Let me see if I can add a bit to the discussion about the ban treaty and customary international law (CIL). The process for creating new rules of CIL is opaque and imprecise, but – bottom line up front – there is (IMHO) virtually no chance that this treaty process, even if it were wildly successful, wildly soon – would be converted into CIL that would have any impact on the US or the other P5.

    The standard way of talking about the creation of CIL is that two elements are required. The “objective” element rests upon a pattern of widespread, longstanding concordant behavior by states. The pattern is not required to be 100% unanimous, and in “modern” CIL, the pattern need not extend for centuries or even decades – but the more states behave in a particular fashion, the better. As one previous commentator has noted, states are not treated identically for this purpose – those that are especially active in the particular field are weighed more heavily. It’s not always easy to identify the specially-affected states (here, would it be the states that possess nuclear weapons; the states that might possess; the states that have had nuclear weapons used or threatened against them; or, given the peculiar power of nuclear weapons, would it be all states?)

    The second element – more problematic here – is the psychological element, asking why the states have behaved according to the observed pattern. In particular, the “opinio juris” element requires that states adopt the behavior out of a belief that it is already legally obligatory. It has to be a behavior that they deem mandatory, not adopted out of habit, courtesy, politics, or strategy. (There is a bit of cumbersome circularity to this element – something becomes CIL only when states (mistakenly) believe that it is already CIL; that has routinely been a hard question on the exams I give in International Law I.)

    So for the ban treaty to become CIL, not only would many/most states have to refrain from possessing nuclear weapons, and not only would that group of states have to include the states that have devoted so much treasure and expertise to procuring nuclear weapons, but in addition, those states would have to believe that the ban is already (or, perhaps, is becoming) legally obligatory even for states that have not joined the treaty.

    The magic of CIL does not happen often in the realm of arms control, but it does sometimes occur. Chemical weapons would be the best example – although we do not have anything like empirical proof of the proposition, most observers assess that the prohibition against use of CW became CIL even before the CWC was in place, and that it therefore binds even states that have not joined that treaty. The prohibition against placing WMD in orbit might be another example – even before (or perhaps concurrently with) the conclusion of the Outer Space Treaty, that restriction was probably accepted by sufficient states, and was motivated by the necessary sense of obligation, that it became CIL.

    In addition (again, as noted by an earlier commentator), a state may exempt itself from an emerging CIL norm by positioning itself as a consistent dissenter or persistent objector against the norm. Again, there are few clear rules, and few manifest examples of this practice, but the standard international law rule would be that if a state is overt and consistent in its rejection of the emerging CIL, it is not bound by it. (If you like, you can see this rule as reversing the “default setting” for state practice regarding treaties: For a treaty, a state is NOT bound, unless it affirmatively opts in, by joining the treaty; for a CIL rule, a state presumptively IS bound, unless it acts overtly to escape.)

    So for the ban treaty, even if, say, 180 NNWS joined the agreement and emplaced new rules among themselves, that widespread concordant behavior would not be sufficient to establish a norm of CIL – states would have to believe, in addition, that the prohibition against possession was already (or was becoming) legally obligatory even against the non-participants, even though those states were the only ones to possess nuclear weapons. And even if the rule did, somehow, evolve into a CIL, it would not be binding on the conspicuous dissenters.

    David Koplow