Text Only | Disarmament Diplomacy | Disarmament Documentation | ACRONYM Reports
Back to the Acronym home page
British Policy
South Asia
About Acronym

Disarmament Diplomacy

Issue No. 42, December 1999

Withdrawal Clauses In Disarmament Treaties: A Questionable Logic?
By Nicholas A. Sims


This article considers the place of the withdrawal clause in disarmament treaties and the question of whether they have been too unthinkingly assimilated into the familiar practice of disarmament diplomacy.

The logic of withdrawal clauses can be accommodated within arms control, but sits less easily with disarmament. There are logical difficulties with the insertion of withdrawal clauses into certain models of disarmament treaty which try to build irreversibility into their structures. More generally, the presence of a withdrawal clause may detract from the expectations of durability which a disarmament treaty regime seeks to promote; it runs counter to the emergence of a 'regime of permanence', which is evolving both in the Biological and Toxin Weapons Convention (BWC) and the Chemical Weapons Convention (CWC).

The conclusion is reached that, as the abolitionist agenda advances, careful thought needs to be given to whether any proposed Nuclear Weapons Convention (NWC) should include a withdrawal clause or not.


It seems certain that the Protocol to the BWC being negotiated by the Ad Hoc Group in Geneva will include a withdrawal clause, along the lines of the one found in the BWC itself at Article 13.2: "Each State Party to this Convention shall in exercising its national sovereignty have the right to withdraw from the Convention if it decides that extraordinary events, related to the subject-matter of the Convention, have jeopardised the supreme interests of its country. It shall give notice of such withdrawal to all other States Parties to the Convention and to the United Nations Security Council three months in advance. Such notice shall include a statement of the extraordinary events it regards as having jeopardised its supreme interests."

The Ad Hoc Group has still to decide the length of notice and the list of bodies to be notified. But the principle of a withdrawal clause based on Article 13.2 of the BWC is not contested.

The BWC was the fifth multilateral treaty in the

field of arms control and disarmament to contain such a withdrawal clause. It followed identical provisions in the NPT (Article 10.1) and the 1971 Sea Bed Treaty (Article 8), which were its immediate predecessors in the Geneva negotiating body. The 1963 PTBT (Article 4) and the 1967 Treaty of Tlatelolco (Article 30) also referred to "supreme interests" and required three months' notice but differed slightly in their wording, especially with respect to the lists of those to be notified.

The inclusion of a withdrawal clause in any treaty may be criticised as offering too easy an escape route once obligations become onerous. It puts temptation in the way of any government which finds a treaty even mildly inconvenient. And it calls into question the lofty notions of 'good faith' and pacta sunt servanda ('treaties are to be kept') in which treaty-makers not unreasonably like to immerse the fruits of their negotiations.

On the other hand, as noted by Alan James in his 1972 analysis of the five treaties of 1963-1972:

"At first sight this arrangement might seem rather odd, and even, perhaps, destructive of the treaties' central purposes. But the ability to shed an obligation lawfully by no means makes one as free, effectively, as if one had no obligation in the first place. And it is arguable that the break clauses make no substantial difference at all to the treaties' strength, being simply an unusually frank acknowledgement of a real possibility and an attempt to minimise the damaging side-effects of such an occurrence."1

Any credible analysis has to embrace both law and politics, and ask the question: "What legal and political purposes does the withdrawal clause serve?"

Legal Purposes of Withdrawal

If it serves a purpose at all in legal terms it is principally that it constrains the otherwise potentially destructive effect on a treaty regime of the international legal doctrine of Fundamental Change of Circumstances. This doctrine is sometimes known as Rebus Sic Stantibus, from the clausula rebus sic stantibus (the little clause 'things remaining as they are'), which is sometimes controversially said to be an implied term in any international agreement.

The withdrawal clause says that not every fundamental change, however defined, suffices as a pretext for withdrawal from these treaties. Instead it introduces three tests:

a. Extra-ordinariness;

b. Treaty relevance; and

c. Magnitude of threat to national security.

These tests, applied in combination, narrow the range of events which can properly be invoked to justify withdrawal.

Only if a state party decides that a particular event is (a) extraordinary, and (b) related to the subject-matter of the treaty in question, and (c) of such magnitude as an actual threat to national security that it has already jeopardised the supreme interests of its country, can it draw up the required statement of explanation and embark on the withdrawal procedure laid down in the treaty.

So far, so satisfactory. The occasions which might precipitate withdrawal are duly restricted by the requirement that each of three tests must be satisfied.

Nevertheless, the withdrawal clause suffers from the disadvantage that the whole assessment is within the sole prerogative of the withdrawing state. It decides for itself whether the three conditions have been met. It exercises its own judgement, which is then the final authority in the matter, regardless of how partial or faulty that judgement appears to others. There is no trace of deference to the judgement of international society, or even to that of the collectivity of states parties, which might be supposed to have some legal standing in the case.

Another criticism of this legal justification for the withdrawal clause is that it assumes a greater legitimacy for Fundamental Change of Circumstances than that questionable doctrine deserves. It undermines, even contradicts, the principle of pacta sunt servanda, cornerstone of the whole law of treaties.2

Length of Notice of Withdrawal3

A subsidiary legal purpose of the withdrawal clause is to specify the procedure for withdrawal, ensuring that all interested parties are informed well in advance, with an explanation. All five treaties in the 1963-1972 series required three months' notice. Walter Krutzsch and Ralf Trapp have written of the equivalent 90 days' notice requirement in Article 16.2 of the 1993 CWC: "The time span should enable States Parties either to try to convince the State Party in question not to insist on its decision (for instance by providing additional assistance against use or threat of use of chemical weapons) or to prepare individually

or collectively for measures adequate to cope with an adverse situation resulting from such a withdrawal."4

For its next treaty, however, the CD doubled the notice required. The 1996 CTBT, in Article 9.3, sets a minimum of six months. The same length of notice is under consideration for Article 15.2 of the BWC Protocol.5

Political Purposes of Withdrawal

The withdrawal clause in multilateral treaties for arms control (but less certainly, as will be argued below, in disarmament treaties) serves two main political purposes. Arguably, these are only the international and domestic-political aspects of the same purpose: the preservation of national security.

Internationally, it offers a legitimate way out if an action by a non-party - testing or acquiring nuclear weapons, for example - fundamentally upsets the whole set of security assumptions upon which the bargain was based. Action by a party to the treaty might also have this effect, but is in principle less likely, and in any case is subject to whatever implicit or explicit sanctions the treaty may contain: sanctions which can logically not be applied to a state outside that treaty.

Domestically, it enables the proponents of a treaty within governments to reassure nervous legislators or military planners that they are not putting at risk the supreme national interests of their country. Without a withdrawal clause, however unlikely it is ever to be invoked in practice, a treaty might fall at the hurdle of domestic political acceptability. But with such a clause a US president, for example, is in a marginally stronger position to obtain advice and consent to ratification from a two-thirds majority in the Senate. Whatever option is being foregone under the treaty in question it is not being given up irrevocably.

Rebecca Johnson has emphasised this safety-net function - and also its limited effect beyond psychological reassurance - in her analysis of the conclusion of the CTBT negotiations in 1996: "The general view in Geneva was that such an interpretation of supreme interests acts as an insurance net (and appeases military factions or reassures hawks during ratification debates). In practice it would be difficult to withdraw on that basis after the treaty had entered into force, for fear of unravelling not only the CTBT, but the wider credibility of international non-proliferation and arms control treaties."6

In those negotiations, moreover, governments had expressed themselves with unaccustomed frankness on the subject of what would trigger the use of the withdrawal clause.

Arms Control and Disarmament

Arms control agreements are security bargains aimed at stabilising a set of politico-military relationships, global or regional, multilateral or bilateral as the case may be. They may include measures of restraint on the testing, manufacture, possession, deployment or use of weapons; but they do not necessarily imply any reduction of armament levels.7

Disarmament, on the other hand, does imply at least a reduction in armament levels and, classically, the renunciation of a type or types of weapons, with reduction to zero. The BWC and CWC do not just seek to preclude any proliferation of the prohibited weapons to new possessors: equally they require the destruction of those weapons by states already in possession of them. An NWC would resemble the BWC and CWC in that respect. In all three cases, abolition is the essence of disarmament.8

That is why the withdrawal clause, which can be accommodated within arms control, sits less easily with disarmament. Abolition is a total, once-and-for-all action by the society of states; and the members of that society can hardly unite with confidence in renouncing the totality of a class of weapons - biological, chemical or, prospectively, nuclear - if there is a legitimised escape route which each member knows each other member is free to take. "The logic of withdrawal clauses implies at least a partial reversibility of the treaties in which they are incorporated. Now it is by no means generally agreed that a disarmament treaty should be reversible."9

The problem is that governments have seldom taken any notice of the academic distinction between arms control and disarmament, and have been notably reluctant to recognise the radical logic of disarmament. Even when congratulating themselves upon concluding the BWC as "the world's first disarmament treaty" - a phrase much used in 197210 - those same governments had only recently restored to its text a withdrawal clause discarded at one stage in 1971.11 They managed to do without one in the next product of the CCD at Geneva: ironically a treaty which belonged to the realm of arms control rather than disarmament. This was the Convention on the Prohibition of the Use of Environmental Modification Techniques for Military or Other Hostile Purposes, which was opened for signature in 1977 and entered into force in 1978. Its Article 7 provided simply for unlimited duration. But when they resumed negotiation of the much more important CWC - important not least because it carried forward the disarmament process from the BWC - they continued to ignore the plea that the error committed at a late stage in negotiating the BWC should not be repeated in the CWC12. The withdrawal clause went in, and stayed in. It was not even a matter of controversy, over the almost 21 years13 that the CWC was in the making at Geneva.

The BWC and CWC

The withdrawal clauses sit uneasily in the BWC and CWC as both treaties develop expectations of durability.

In the BWC a 'regime of permanence' has grown up, through the first four review conferences and their final declarations, notably through the renunciation of those reservations to the 1925 Geneva Protocol which reserved a right of retaliation in kind. The logic of biological disarmament requires an absolute renunciation of biological warfare.

Other elements in the evolving BWC 'regime of permanence' include the perpetuation of a five-year cycle of review conferences, the extension of its international obligations into domestic legislation (patchily achieved) and, in prospect, for States Parties to the new BWC Protocol, a permanent Organisation for the Prohibition of Biological and Toxin Weapons.

The CWC already possesses, established from the outset, its permanent Organisation for the Prohibition of Chemical Weapons (OPCW). It includes a clear-cut prohibition of chemical weapon use in its Article 1 (which also logically requires the abandonment of the precaution of reserving a retaliatory option), together with the obligation to entrench national implementation in domestic legislation and the establishment or designation of CWC National Authorities.

All these elements contribute to the evolution of a 'regime of permanence' for the CWC as well, deepening its expectations of durability.

This is not to pretend that all is well with the two conventions. Far from it: both suffer from lack of universal participation, suspicions of non-compliance by some States Parties and fears of proliferatory ambitions on the part of non-States Parties, not to mention North-South tensions over issues of free trade and export controls and persistent anxieties over terrorism, the verifiability of the conventions and their capacity to deter violations through enabling timely detection and reaction.

But what is significant is the response to these many concerns. There is a wide range of responses, from criminalising the actions of individuals and terrorist organisations in international law to building stronger treaty institutions, and embracing the whole array of reinforcement measures that the Ad Hoc Group aims to put into the Protocol in order to strengthen the BWC (see Henrietta Wilson's report on the BWC negotiations in this issue).

None of these responses includes dismantling the conventions. On the contrary, the emphasis is all on universalising them and making them work more effectively in the service of an ever-widening collectivity of States Parties. Indefinite duration of ever more deeply entrenched and irreversible treaty regimes of biological and chemical disarmament is the assumption on which all this activity rests.

Withdrawal clauses add nothing. If anything, they put a question mark over the assumption of indefinite duration and run counter to the emergence of a 'regime of permanence' for the BWC and CWC.

GCD & a Nuclear Weapons Convention

Some treaties build irreversibility into their structure, emphasising a 'regime of permanence' from the beginning in their conceptualisation.

Reversibility can never be ruled out in practice, but it is illogical to write such a possibility into a treaty for General and Complete Disarmament (GCD). Indeed it was when GCD was last a live issue on the diplomatic agenda, roughly from 1958 to 1964, that the question of reversibility arose.

It was answered in part by giving the concept of GCD a treaty structure which emphasised sequence: states would only proceed to the next stage if satisfied that the first stage had been successfully completed. What should be disarmed at which stage, supranational control, and the verification of baselines and remainders, proved to be the major areas of disagreement in rival draft treaties over which the GCD project foundered.14

And what of an NWC? The assumption is generally made that its content, like its title, would broadly follow the pattern of the BWC and CWC. Indeed, much of its appeal resides in the prospect of an NWC completing the disarmament of weapons of mass destruction (WMD), which the treaty-making of 1972 and 1993 began.

If, however, it unthinkingly follows the pattern of the BWC and CWC (or that of the NPT and test ban treaties) in respect of a withdrawal clause, it will be an opportunity missed. There may be a case for a withdrawal clause in certain models of an NWC. But, if there is, it is one that has to be argued afresh rather than relying on precedents in the nuclear arms control treaties, which did not aspire to disarmament, or the biological and chemical disarmament treaties, which did, but with a questionable logic when it came to withdrawal clauses.

Interestingly, the Model Convention on Nuclear Weapons produced by the International Association of Lawyers Against Nuclear Arms (IALANA), International Network of Engineers & Scientists Against Proliferation (INESAP) and International Physicians for the Prevention of Nuclear War (IPPNW) deliberately prohibits withdrawal, in its Article 18.5, and explains that this reflects "the view that the prohibition of nuclear weapons, and the obligation to eliminate them, have entered the realm of customary international law from which there should be no exception."15


Abolitionists have to confront the objections that weapons cannot be disinvented or knowledge disowned. But these objections are typically countered by strengthening treaty provisions of verification and control, so that the future regime of nuclear disarmament is rendered more resilient and generates confidence in its permanent duration; by devising robust international institutions; and, not least, by emphasising the importance of popular support so that the passionate commitment to a world free of nuclear weapons would become deeply embedded in civil society the world over.

Amendment and review procedures should enable negotiators to build in sufficient flexibility for disarmament treaties to survive. Moreover, it is in negotiating tough requirements for entry into force that states can best insure against the sort of fundamental disturbance of security assumptions that might otherwise precipitate a future withdrawal from the treaty.

As the abolitionist agenda advances, an NWC is becoming ever more conceivable - unlike GCD, which has remained moribund since 1964. 'Getting to Zero' and its attendant problems of verification and security have attracted much interest.16 Careful thought needs to be given to whether a withdrawal clause would strengthen or weaken particular models of treaty-making for permanent nuclear disarmament.

Notes and references:

1. "Law and Order in International Society", Alan James, The Bases of International Order: Essays in Honour of C A W Manning (London: Oxford University Press, 1973) p78.

2. "Rebus Sic Stantibus", in Edmund Jan Osmanczyk. The Encyclopedia of the UN and International Agreements, 2nd edn (London: Taylor & Francis, 1990), pp 739-740.

3. Outside the field of arms control and disarmament, withdrawal clauses, where they exist at all, commonly specify a longer period of notice: one year is usual, and two years not unknown. It is rare for any statement of reason to be required, let alone any test of justification. Where there is no withdrawal clause at all, the effect of Article 56 of the 1969 Vienna Convention on the Law of Treaties is to impose a minimum notice period of "not less than twelve months" and to place the burden of proof on the state wishing to withdraw. It has to establish "that the parties intended to admit the possibility of denunciation or withdrawal" or, alternatively, that "a right of denunciation or withdrawal may be implied by the nature of the treaty": a tough test.

4."A Commentary on the Chemical Weapons Convention", Walter Krutzsch & Ralf Trapp (Dordrecht: Martinus Nijhoff, 1994), p248.

5. "Article XV: Duration and Withdrawal", Nicholas A Sims, Evaluation Paper No 4 in Graham S Pearson & Malcolm R Dando (series eds), The BWC Protocol: Evaluation Papers (Bradford: Department of Peace Studies, University of Bradford, 1999) pp 4-5.

6. "A Comprehensive Test Ban Treaty: Signed but not Sealed. A Review of the CTBT Negotiations in the Conference on Disarmament January-September 1996", Rebecca Johnson, Acronym Paper No 10 (London: Disarmament Intelligence Review, May 1997) p60.

7. This description of arms control is based on the classic definition by Hedley Bull, The Control of the Arms Race (London: Weidenfeld & Nicolson, 1961), introduction, p ix.

8. A Nuclear Weapons Convention has been associated particularly with the Malaysian resolution adopted every year since 1996 by the UN General Assembly, but is not excluded (as a more distant eventuality) by the more 'moderate' resolutions on nuclear weapons associated with Japan and (since 1998) with the New Agenda Coalition. On the significance of these resolutions, their varying majorities and oppositions, and the First Committee debates which have revolved around them, see "First Committee Report", Rebecca Johnson, Disarmament Diplomacy No 10 (November 1996) ; No 20 (November 1997); No 32 (November 1998); No 41 (December 1999). There has also been much NGO interest (see notes 15 and 16), including a Model Nuclear Weapons Convention which, in its 1999 form, was a revised version of the 1997 draft submitted by Costa Rica in UN Doc A/C.1/52/7.

9."Approaches to Disarmament", Nicholas A Sims (London: Friends Peace & International Relations Committee, 1974) p30.

10. Sims (1974) pp 38-39.

11. "Biological disarmament: Britain's new posture", Nicholas Sims, New Scientist, vol 52 pp 18-20 (December 2, 1971).

12. Sims (1974) p30; repeated in the revised edition (1979) p 52.

13. From March 1972 to August 1992, i.e. 21 annual sessions of the Conference on Disarmament and its predecessors.

14. "Approaches to Disarmament", Nicholas A Sims, revised edition (London: Quaker Peace & Service, 1979) pp 104-109; "The Forgotten Treaties: A Practical Plan for World Disarmament", Allan McKnight & Keith Suter (Melbourne: Law Council of Australia, 1983) pp 16-54.

15. "Security and Survival: The Case for a Nuclear Weapons Convention", International Association of Lawyers against Nuclear Arms, International Network of Engineers and Scientists Against Proliferation, International Physicians for the Prevention of Nuclear War (Cambridge, Mass.: IPPNW, 1999) section 2 p53. It should be noted that a square-bracketed addition to Article 18.5 would bring this prohibition of withdrawal into force only "upon ratification by all Nuclear Weapons States".

16. See "Nuclear Weapons: The Road to Zero", [Sir] Joseph Rotblat (ed), Oxford: Westview Press for Pugwash, 1998; "Laying the Foundations for Getting to Zero: Verifying the Transition to Low Levels of Nuclear Weapons", Patricia M. Lewis, VERTIC Research Report No 1, September 1998; "Verifying the Transition from Low Levels of Nuclear Weapons to Zero", Tom Milne & Henrietta Wilson, VERTIC Research Report No 2, June 1999; "Virtual Nuclear Capabilities and Deterrence in a World Without Nuclear Weapons", George Paloczi-Horvath, VERTIC Research Report No 3, October 1998; "Sustaining a Verification Regime in a Nuclear Weapon-Free World", Suzanna van Moyland, VERTIC Research Report No 4, June 1999.

Nicholas Sims is a Senior Lecturer in the Department of International Relations at the London School of Economics and Political Science, University of London. He specialises in the diplomacy of disarmament and is the author of "International Organization for Chemical Disarmament"(1987) and "The Diplomacy of Biological Disarmament" (1988).

© 2000 The Acronym Institute.

Return to top of page

Return to List of Contents

Return to Acronym Main Page