Reflections on Nuclear Transparency and Irreversibility: the re-regulation of partially disarmed states

Introduction

Irreversibility has become an important objective in the post-Cold War environment, reflecting the double determination that there should be no backsliding towards the confrontations and repressions that have scarred this Century, and that recent economic and political gains should be consolidated. Transparency has come to be regarded as one of the main instruments for achieving this irreversibility and for bringing about improvements in governance.

The term 'transparency and irreversibility' therefore has an appealing ring to it. It has long had special relevance to nuclear arms control. The non-proliferation regime can, for example, be described as an institution for sustaining the irreversibility of nuclear renunciations through, amongst other things, transparency measures (notably safeguards). Recently, the ambition to strengthen the irreversibility of arms reductions was recognised in the joint US-Russian statement issued in Helsinki. The further that the international community moves in the direction of nuclear disarmament, the more imperative will it become to build confidence in irreversibility, and the greater the burden that will be placed upon transparency. That effort must focus in substantial part on the control of fissile materials.

Yet transparency and irreversibility are complex notions, as will be apparent in the first part of this paper. Their application and achievement are especially complex - yet essential - in the context of the nuclear weapon states. My main point is that confidence in the irreversibility of deep arms reductions, and in the water-tightness of fissile material stocks and infrastructures, cannot be achieved without advances in transparency and verification; but that these advances require more extensive regulatory changes than have been attempted or even envisaged so far. How should a partially disarmed state be regulated? This is the fundamental question that needs answering if, as most of us would probably agree, complete disarmament is still far away. Besides establishing a new international norm, the great importance of the Fissile Material Cut-off Treaty (FMCT) negotiations lies, in my view, in the impulse that they will give to advances in legal and regulatory processes. But the FMCT cannot carry this burden alone: fissile material stocks, and issues of principle, also need to be addressed.

On irreversibility

Irreversibility is desired when a condition or state of affairs is reached that is judged to be superior to some other (often prior) condition or state of affairs. ₁ The corollary is that reversibility is desired when people or institutions find themselves in a condition that they judge to be inferior. To give the obvious example, the irreversibility of peace and reversibility of war, and the irreversibility and reversibility of factors contributing towards them, are perennial concerns in international politics.

Once the superior condition is identified, there are in at least four 'objects' that states and other actors try to render irreversible:

1. The condition itself: to render irreversible the superior condition when it has been achieved in part or in whole. Thus the international community tries to hold states to their renunciations of nuclear weapons once they have declared them.

2. Movement: to render irreversible the predisposition and movement towards a superior condition where it has not yet been sufficiently achieved. Thus efforts are made to maintain trends towards universal membership of and compliance with the NPT, and to define new measures (such as nuclear-weapon-free zones) that will bring the international community within closer reach of the final goal. Note that confidence in the eventual achievement of superior conditions depends upon the irreversibility of trends towards them, and upon the reversibility of trends away from them (including actions on the margins, such as North Korea's weapon programme, that threaten to trigger more widespread backsliding).

3. Measures: to render irreversible the integrity and effectiveness of particular measures and institutions that aim to sustain superior conditions and the trends towards them, and/or that achieve the reversibility of inferior conditions and the dynamics underpinning them. The CTBT and its verifying institutions provide an example: they help guarantee the reversal of nuclear arms races by enshrining in international law and practice the irreversibility of the cessation of explosive testing.

4. Processes: to render irreversible the processes that sustain superior conditions, the trends towards them and the measures that support them. Examples would be the various UN instruments of mediation, negotiation and verification; or more fundamentally, the processes of international law.

Regimes may be described as institutions for defining and achieving superior conditions, and for rendering those conditions and the movement towards them irreversible. We should recognize, however, that regimes are unlikely to achieve 'permanent irreversibility' on their own. Usually, new conditions only acquire true permanence when profound systemic changes and/or changes in values have also taken place. ₂ A fundamental uncertainty concerning nuclear weapons is whether observed systemic changes (notably in great power relations, in economic and political interdependence, and in military capabilities and strategies), together with changes in values (notably in attitudes towards war as a political instrument, and towards the usage or threatened usage of genocidal weapons), are sufficiently strong to bring about their permanent obsolescence. If they are not so strong, multilateral regimes will have to continue carrying a heavy burden in dissuading states from acquiring and using nuclear weapons. The non-proliferation regime will remain vulnerable to breakdown.

For almost half a century after the discovery of nuclear fission, two processes were in evidence. One was the process of nuclear armament, the other the process of nuclear renunciation and disarmament. The objective throughout this period was to achieve the reversibility of the former and the irreversibility of the latter, the ultimate aim being to arrive at a universal and irreversible 'superior condition' - complete nuclear disarmament. The non-proliferation regime embraced all four 'objects' of irreversibility described above: the superior condition, and the movement, steps and processes leading towards it. It has been spectacularly successful in bringing the great majority of nation states (183 at the latest count?) to the 'superior condition' of disarmament, and in providing confidence that they will stay in that condition. However, the regime was initially designed to achieve neither the reversibility of nuclear armament, nor the transformation of nuclear weapon states into non-nuclear weapon states. Indeed, the NPT is deeply ambiguous on this count: it grants the five nation states that qualify for the title 'nuclear weapon state' irreversible legal rights to develop, possess and use nuclear weapons; simultaneously, it places them under the irreversible obligation to negotiate away those weapons. This ambiguity lies, of course, at the heart of contemporary debates about the regime.

Since 1986, a process of disarmament has, nevertheless, been under way in the NWS, and especially in the USA and former Soviet Union which assembled by far the largest nuclear arsenals during the Cold War. However, it differs from the process of disarmament experienced by other states in three fundamental respects:

1. It involves the dismantlement of a huge quantity and variety of weaponry and componentry, along with the dismantlement or conversion of capacities assembled for producing them. Besides the scale of the task, these artifacts and capacities embody knowledge of nuclear weapon technology so that care has to be taken that it is not dispersed into the hands of other actors. Still more important, disarmament for these states involves the dismantlement or adaptation of institutions and cultures that have put down deep roots within their societies.

2. The 'superior condition' towards which this disarmament process is moving has not been clearly defined. Is the desired movement towards complete disarmament, or lower levels of armament, or small reserve holdings of nuclear weaponry, or what? I am not alone in arguing that the ultimate objective probably cannot be defined at present, and that it would even be imprudent to attempt to do so: the disarmament of heavily armed nuclear states, and the ending of dependence on nuclear deterrence, is bound to be an evolutionary process that can only proceed step by step and will remain comparatively open-ended until the final destination is in sight. ₃ Indeed, one can argue that any other process would be unsafe and inefficient. However, this begs important questions about the energy and directedness of evolutionary processes, and about the degree to which consensus can be built around them.

3. There are no pre-defined and mandatory requirements for submitting the disarmed materials, components and facilities to multilateral verification. Whereas states which join the NPT as NNWS have to follow a prescribed regulatory path (notably involving IAEA safeguards), the same does not apply to NWS engaged in acts of disarmament. In the main, the regulations applied by and to them have so far been adopted either unilaterally or through bilateral negotiation: they are intrinsically voluntary and discretionary.

As a consequence, we face not one but two confusions. The first is the confusion over the final destination - the 'superior condition' that nations with nuclear arms should be heading towards. The second is the confusion over the regulatory objectives and obligations of partially disarmed weapon states. If complete disarmament is not on the cards, or is unavoidably a long-term prospect, how should partially disarmed states be regulated in the meantime?

This latter problem is fundamental where fissile materials are concerned, as will be discussed below. It is problematic partly because the major part of the military nuclear 'estates' in Russia and the USA is being disarmed, so that a small minority will soon remain tied to operational nuclear weaponry. In previous times, the regulatory stance of nuclear weapon states was driven largely from the majority military domain. But should that continue to be the case if the military nuclear estate is now so diminished? Should the tail be allowed to wag the dog? To continue the metaphor, is the regulatory dog bound to be a mongrel, and if so, what kind of mongrel would be both effective and acceptable given the high standards of pedigree demanded elsewhere?

On transparencyIn international politics, transparency is sought in regard to the intentions, capabilities and decision-making processes of nation states. Transparency of intentions is arguably the most important amongst them, although it is the one that states may be keenest to disguise. Transparency of capabilities is often given priority by states partly (only partly) because it gives clues to other states' intentions.

In any setting, transparency usually fulfils four purposes, in no particular order of importance:

1. Administrative effectiveness and efficiency. It discourages sloppy practices, encourages the spread of best practices, and through standardisation diminishes transaction costs.

2. Political accountability. Transparency provides some protection against the abuse of power, and may improve the quality of decision-making.

3. Confidence in good behaviour. Transparency encourages trust amongst states, and between states and peoples. Furthermore, transparency discourages actions that may enhance mistrust - it has a deterrent effect - and provides warning of unwelcome acts or developments.

4. Analytical abilities. Transparency improves decision-making and administration insofar as they depend upon analysis of observable trends, circumstances and histories. It is also assists 'problem identification': if, for instance, pollutants in rivers are not measured, the size, nature and even existence of problems caused by them cannot be assessed.

Transparency is not invariably desirable. We can all recognize circumstances, in our own lives as well as in relations between states, when greater transparency would be harmful. But why is opaqueness often deliberately fostered in preference to transparency? There are at least three reasons:

1. Opaqueness confers advantage upon particular groups in society: it allows them to set agendas that serve their special interests, to preserve autonomy in decision-making, to maximize their power-through-knowledge, and to avoid scrutiny by competitors or publics.

2. Public welfare may in some circumstances be best protected by opaqueness, or at least by the internalisation of transparency. Was Orson Welles right to inform the inhabitants of San Francisco that Martians had landed? More prosaically, the protection of individuals, installations or possessions against criminal or terrorist attack requires a degree of opaqueness. We do not, for instance, publicize the contents of our homes.

3. Opaqueness is endemic to systemic relations in certain contexts. It was endemic to nuclear strategy as practised during the Cold War. Belief in the deterrent effect of nuclear weapons - and the quest for strategic advantage - depended upon maintaining uncertainty about intentions and capabilities, even if a degree of transparency was sought through arms control measures to reduce uncertainty where it threatened a serious destabilization of strategic relations. One should add that the elites responsible for nuclear decision-making had a particular interest in sustaining opaqueness within their national polities given the controversies that surrounded nuclear weaponry.

The important point here is that processes of nuclear disarmament carry a predisposition towards transparency, whereas processes of nuclear armament carry a predisposition towards opaqueness (albeit moderated by arms control). Indeed, transparency became from the outset a primary instrument of irreversibility within the nuclear non-proliferation regime, while nuclear weapon states retained freedom from transparency in the presumed interest of deterrence. These contrasting standards of transparency became embedded in the NPT's text. But this dichotomy has lost much of its political and functional legitimacy following the end of the Cold War, the great arms reduction that came in its wake, and the disintegration of the Soviet political system. As we shall see below, the appropriate standards of transparency for partially disarmed states are now central issues in the development of nuclear arms control and disarmament policies.

Finally, it should be noted that states seeking to acquire nuclear weapons have also constantly sought to maximise opaqueness. In their cases, foreign powers have tried to impel greater transparency, and to gain more information about their activities by various overt and covert means. Within the framework of non-proliferation policy, transparency has value both as an instrument for sustaining irreversibility (in regard to NNWS parties to the NPT) and as an instrument for encouraging and preparing the ground for reversibility (in regard to 'proliferants').

A caveat should be added to the above observations about transparency. While opaqueness was given priority by the USA and Soviet Union during the Cold War, the deterrent would not have deterred if the opponent had been kept completely in the dark. Total opaqueness would have been counterproductive. As a result, each side engaged unilaterally in 'controlled transparency'; but this transparency was often bent in order to deceive (dishonesty was expected, and disinformation was raised to an art), and it was always tightly controlled and manipulated by the individual state. Honesty was only sought by arms controllers, and even for them it was seldom absolute.

In the contexts of non-proliferation and disarmament policy, transparency is also carefully controlled by states, but with two important differences: dishonesty is unacceptable; and there is no room for discretion - states are required to provide certain kinds of information according to agreed rules and procedures, from which there can be no departure.

On transparency, irreversibility and fissile materials

Within the non-proliferation regime, fissile materials have long carried much of the burden of attaining irreversibility. Confidence that states and other actors are not acquiring nuclear weapons or weapon capabilities is provided by monitoring material stocks and flows, i.e. through transparency and verification measures known as international safeguards. The reasons for focusing on plutonium and enriched uranium are that they are the only materials from which viable nuclear weapons can be manufactured (there are no substitutes), they are difficult and costly to produce, they have special physical qualities that allow them to be measured and detected, and they are prevalent in both civil and military activities.

Hence the provisions in Article III of the NPT, elaborated in the model safeguards document INFCIRC/153, that define the standards of transparency and intrusiveness with which non-nuclear weapon states Parties must comply. Article III operates in conjunction with Article II which contains the obligation not to acquire nuclear weapons. Together, we should note, they oblige the NNWS to accept a preemptive and verified 'cutoff' on the production of fissile material for weapons purposes. In essence, the IAEA safeguard system, as applied in these NNWS, drew a line around the materials and facilities used for civil purposes and provided reassurance that there would be no passage ('diversion') across that line into military activity. The whole nuclear domain in the NNWS was thus opened to verification. As we now know after the Iraqi débacle, the standards of transparency and intrusiveness applied by the IAEA were insufficient. They have been redefined in the Protocol that was agreed in Vienna in May 1997 and which provides added layers of reassurance that irreversibility means irreversibility for the NNWS.

We should note in parenthesis that the transparency entailed by safeguards is heavily rule-bound ('controlled') even under the new Protocol. The flow of information about fissile materials is from a division within a department appointed by government to the IAEA's safeguards division. It cannot be distributed outside this specific set of institutions. Furthermore, transparency does not extend to the publication of safeguards information except in highly aggregated form. The main transparency 'contract' is therefore between governments and international agencies, rather than between governmental executives and parliamentary institutions.

Article III of the NPT contains no reference to the nuclear weapon states, which are under no treaty obligation to submit any of their activities to international safeguards. Furthermore, the NPT gives the NWS carte blanche to produce as much material for weapons purposes as they like - there are no 'cutoff' provisions matching those arising from Article II. The only concession made by the five NWS Parties was to sign 'voluntary offer' agreements with the IAEA in the 1970s and 1980s which allow the Agency to inspect certain civil facilities placed unilaterally upon 'facilities lists'. The 'Timbs criteria' were drawn up to guide the IAEA's selection of facilities for inspection, the main criterion being that safeguarding a given facility should help the IAEA to perform its primary safeguards duties in the NNWS, notably through the experience gained in safeguarding certain kinds of fuel-cycle facility. Together with the financial constraints on the IAEA's budget, this resulted in a severe rationing of inspection time in the NWS.

The verification of fissile material production, stocks and usage in the NWS was therefore largely discretionary and limited only to selected civil facilities and the materials within them. ₄ It was also reversible: facilities and materials could be withdrawn from safeguards. In addition, inventories of fissile material could be expanded without restriction in both civil and military domains. The NWS therefore retained rights under the NPT to draw their own lines between what could and could not be safeguarded, a line that was movable and porous; and they retained freedom to produce for any purpose on any scale. There was no regulatory concept of 'excess'.

This was the NPT/IAEA system. The Euratom system was, however, different. Under the Euratom Treaty, all members of the European Union are required to submit 'materials declared to be civil' to Euratom safeguards. Britain's and France's civil material holdings, together with facilities involved in their production and storage, are therefore routinely submitted to Euratom inspection. In their cases, a definite line is therefore drawn around military activities which alone are exempted from safeguards (Britain and France also retain freedom to produce). The reason for the discrepancy between the NPT and Euratom systems was that the latter, unlike the former, was built upon the principle of universality: all states within the Union have equal rights to engage in civil nuclear activities, and must be subjected to identical regulations. This principle was waived only in relation to military nuclear activities. In many ways, it is regrettable that the IAEA safeguard system did not imitate the Euratom system.

During the Cold War, the exemptions granted to the NWS under the NPT/IAEA system were justified on the grounds that 'diversion to military purposes' was both legitimate and had no significance in their contexts. Verifying that diversion was not occurring from any purposes to military purposes would therefore be a waste of scarce resources. In today's circumstances, however, it does matter, for at least five reasons:

1. If new boundaries are to be placed around and between material inventories, and the 'rules of the game' are to be changed in regard to at least some of those inventories (through the Cutoff and other measures), there will have to be reassurance that material is not re-crossing the boundaries. Amongst add things, this would increase confidence in compliance with arms reduction agreements.

2. The myth that fissile materials were perfectly safe in the hands of the NWS was exploded by the breakdown of Soviet political and administrative structures, and the revealed inadequacy of regulatory practices in the FSU. In all NWS, there is a finite risk of diversion of materials to non-state actors, to renegade actors within the NWS, and to state actors outside the territories of the NWS.
3. Special vigilance is required given the huge quantities of weapon-grade plutonium and uranium possessed by the NWS, the long time-scales of disposition, and the need to continue storing weapon components prior to dismantlement.

4. Sensitivities to inequities in the regulatory obligations of NWS and NNWS are becoming more acute as nuclear arsenals lose salience and are reduced in size, and as awareness grows of the scale of the 'non-military' inventories and infrastructures that are still exempted from safeguard. These sensitivities were very evident in the recent debates on the universality of 93+2 obligations.

5. Diversion from any source, in any nation state, would be deeply threatening to complete nuclear disarmament. Insofar as expectations of eventual disarmament have strengthened since the end of the Cold War, there is a normative presumption that the NWS should take preparatory steps to enable non-diversion to be comprehensively verified on their territories.

The tasks that lie ahead become clearer if we look, in broad outline, at the inventories held by the NWS. With the exceptions of Britain and France under the Euratom Treaty, there was essentially one inventory of fissile materials in each NWS during the Cold War. It was divided into two categories, depending on whether materials and facilities were declared eligible or ineligible for international safeguards by the NWS in question under its voluntary offer agreement. The eligible category contained most but not all civil materials, while the ineligible category contained all military materials and some civil materials. It was taken as given that all military materials were required for operational military purposes. In the mid-1980s, around 2000 tonnes of fissile material were unequivocally assigned to military purposes in the NWS whilst around 250 tonnes were held in the civil domain, a proportion of which was eligible for IAEA safeguards (in practice, only British and French materials were under routine inspection).

A decade later, there is still essentially one inventory in the NWS, but it can now be divided into three regulatory categories. The first category contains, as before, most but not all civil materials; as before, little is under routine inspection outside Euratom, although most is eligible for inspection. ₅ The second category contains materials that are unequivocally assigned to operational military purposes. The third category contains what may be termed 'transitional' materials: they are no longer required for military purposes, and it is assumed that they are 'in transition' from the military to the civil domain, even if a small fraction has yet to make that transition.

Table 1 provides a breakdown of fissile material holdings by country, including the three de facto nuclear weapon states. Three immediate conclusions can be drawn from it:

1. As indicated, the figures in the Table are no more than 'illustrative estimates'. Given the present state of public and private knowledge, it is not possible to offer 'hard estimates', let alone the actual numbers. Only the United States has published data on its inventories produced historically for military purposes; and as there is no accepted definition of what constitutes 'excess to military requirements'. As a result, the boundary between the 'military' and 'transitional' inventories is discretionary. Note that my estimate of the scale of this 'excess' is much larger than the quantities declared by the NWS which remains highly conservative.

2. However one calculates it, the 'transitional' inventory is extremely large. According to my estimate, it accounts for more than half of the global inventory (military + civil) of fissile material. Nearly all is weapon-grade, and it is contained in a great variety of forms. A tiny proportion has so far been placed under IAEA safeguards (only in the USA), although there are commitments to increase the quantities so safeguarded.

3. There is enormous variety across countries in the scale of inventories, and in their distribution across the three categories. This variety would become even more pronounced if distinctions were drawn between plutonium and HEU. Four differences should be noted in particular: (a) the bulk of the 'transitional' material is held in the USA and Russia; (b) among the NWS, only the USA and Russia have acknowledged that they possess excess stocks; (c) the USA, Russia, France and the UK have large holdings of 'civil' material, but only in the last two countries are they under routine international inspection (due to Euratom, France and the UK alone have well established safeguards cultures); and (d) the quantities in India, Israel and Pakistan are much smaller, albeit politically highly sensitive. As a consequence of these differences, the costs and benefits faced by participants in Cut-off negotiations will be asymmetrical - an unavoidable source of political difficulty and complexity.

   A further distinction is important. Whereas the transitional inventories held by the NWS are in motion (one hopes) from the military to the civil domain, the direction of motion of the Indian, Pakistani and Israeli inventories is ambiguous. Internally, these materials are regarded as 'military', whereas externally there is reluctance to grant them that status. The 'non-proliferation objective' in each case is that the material should be moved into the civil domain, not in the other direction.

In these contexts, what is one trying to render irreversible, and why? There are two specific and one general objectives:

1. Production. To end the production of fissile materials for weapons purposes, thereby capping the unsafeguarded inventories that could be acquired by any country from extant production facilities.

2. Stocks. To render irreversible the transfer of fissile materials out of redundant military inventories into transitional inventories, and from transitional inventories into civilian inventories.

3. Re-regulation. To achieve an irreversible movement towards a regulatory structure in partially-disarmed states that is capable of supporting the above objectives, that is consistent with the regulatory norms and practices established within the non-proliferation regime, and that can assist the longer-term goal of nuclear disarmament.

If we take arms control responsibilities seriously, the achievement of all three of these objectives is indispensible. Except in regard to the de facto weapon states, an end to production would lack functional (if not regulatory) significance if it were not accompanied by constraints on access to stocks which otherwise would provide a 'surrogate production system' that could be turned on or off at will. Similarly, constraints on stocks would have limited significance if not accompanied by a ban on production. And one cannot address either agenda (production or stocks) without significant developments in regulatory norms and practices; but those developments will not come about without negotiation on specific measures. Specific negotiations on production and stocks are therefore needed to drive regulatory improvements.

So these agendas are not optional. The main question is how rather than whether to pursue them.

Many steps have already been taken - or are being taken - that contribute to the three kinds of irreversibility identified above. They include the unilateral announcements by four NWS that they have ceased producing fissile material for weapons purposes; the initiation, albeit the slow initiation, of programmes to dispose of excess HEU and plutonium, thereby erecting technical barriers to re-acquisition; the programmes to enhance physical protection in the FSU; the International Plutonium Management discussions in Vienna which may lead to publication of inventories of non-military holdings of plutonium; the NWS' acceptance, in their endorsement of the NPT Principles and Guidelines, that "fissile material transferred from military use to peaceful nuclear activities should, as soon as practicable, be placed under IAEA safeguards, in the framework of the voluntary offer agreements"; and the trilateral Russia-US-IAEA discussions on verification of materials arising from arms reductions.

All of these steps are very valuable, and are the outcomes of commendable efforts by various individuals, governments and agencies. What is lacking, however, is a treaty framework (or set of frameworks) that can give the three agendas greater coherency and energy. This is why the FMCT is, in my view, so badly needed - as a policy driver which would require governments and other relevant institutions to decide the directions in which the regulatory apparatus in NWS should move, and to act accordingly. However, the FMCT needs to be accompanied by complementary negotiations on fissile material stocks. Elsewhere, I and my co-authors have argued that a treaty on stocks might best be negotiated in parallel with the FMCT, and that it should be initiated by the USA and Russia in consultation with other states and with safeguards agencies. ₆ It is possible that START III might provide one means of driving this process, since the verification of warhead dismantlement has been placed on its agenda.

All of these endeavours would be helped by the development and enunciation of a set of principles that would underpin the re-regulation of partially disarmed states. What might those principles be? Here are some candidates:

• Irreversibility: fissile materials removed from dismantled weapons and redundant stocks may not be reused for weapons purposes. Those materials will be transferred into civil inventories.

• Transparency: full and detailed inventories of material acquisitions and holdings will be established, and a 'presumption to disclose' this information will replace a 'presumption to withhold'.

• Minimization: production of fissile materials for weapons purposes will be ended, and stocks held in military inventories will be minimised - they will be scaled only to meet operational requirements consistent with arms reduction treaties (reserve stocks will not be held). Steps will be taken to eliminate stocks of excess fissile material, in accordance with the spent-fuel standard.

• Equality of verification standards: all redundant materials and facilities will be made available to international verification, with the object of matching wherever possible the standards of verification applied in the non-nuclear weapon states.

• Permanence of verification: materials and facilities may not be withdrawn from safeguards or other forms of verification.

Before drawing to a conclusion, let me elaborate briefly on two of these themes. The first is the principle of minimization. For regulatory purposes, there are two primary issues: one is how to set in concrete, via the Cut-off, the outer boundary around material inventories produced for weapons; the other is how to define the inner boundary between 'military' and 'transitional' inventories. The NWS have been reluctant to engage in any discussion, bilateral or otherwise, of the latter boundary. Yet setting this boundary is central to the issue of irreversibility, especially in countries with large excess stocks. If it remains uncharacterized, large unverified reservoirs of fissile materials and weapon components could be kept in store, awaiting possible reactivation for military purposes.

Why is there such reluctance? One reason is that important constituencies in the NWS are not yet prepared to commit themselves to the irreversibility of arms reductions, and they need to be pampered. Another is that the act of setting this boundary would send ripples of transparency in every direction, with uncertain and possibly annoying consequences for decision-making elites. Still another reason put forward is that the boundary around 'military' inventories must be kept fuzzy (a) to protect information about warhead numbers, and (b) to protect warhead design information if those numbers were disclosed.

The issues of transparency of warhead numbers and material stocks are therefore connected. If the German proposals for a register of warheads were adopted, one of the barriers to (or excuses for) non-disclosure of material inventories would be removed. ₇ But how serious a problem is the linkage between disclosure and the dispersal of design information? Its importance can be exaggerated. Such information may be useful to designers, but how useful? What is its 'marginal utility' to inexperienced designers, especially if only aggregate quantities are being revealed? Given the huge amount of information on weapon design that is already available in the open literature, would it make any difference if the British government, for instance, announced that the UK possessed 100 warheads and required 800 kilograms of plutonium and 2000 kilograms of HEU to sustain them? (I have invented these numbers!)

The second issue is that of material accounting. Too little attention has been given, in my view, to the US Department of Energy's Openness Initiative which has led to a much more detailed accounting of all fissile materials produced and located at DOE sites. Besides the transparency that this has generated, the Openness Initiative drove operators and administrators to engage in a detailed examination of past records and to explain - albeit with incomplete success so far - any materials that were unaccounted for. It has also helped them arrive at a better definition of the clean-up problems faced at DOE sites.

Unfortunately, this Initiative has not been matched by the other NWS. They should be given every encouragement to reciprocate. Indeed, it could be argued that they have a responsibility to follow the US example if they take their commitments to nuclear disarmament seriously. In that event, they would individually be required to present the IAEA with an initial inventory so that full-scope safeguards could be applied. None of the NWS would currently be able to comply with this requirement - even the USA would struggle, especially in regard to HEU inventories. The same observation probable holds true for the de jure weapon states. There is some urgency to launch these accounting initiatives as many of the operators and administrators that assembled, and may still hold, records in the 1950s, 1960s and 1970s are ageing. The importance of involving such personnel was demonstrated in the US initiative, as well as in the preparation and verification of South Africa's initial inventory.

Conclusion

The irreversibility of the shift away from nuclear confrontation involves much more than the control of fissile materials. It will depend, inter alia, on steps to reduce the readiness of nuclear weapons, to implement the nuclear test ban and, most importantly, to develop friendly and durable relations between former adversaries. In the long run, it will depend above all on shifts in ideologies, geo-political structures, military strategies, and international governance.

This said, fissile materials must lie close to the centre of any irreversibility strategy. Let me make two points in conclusion.

The first concerns the confusion over the 'superior condition' that the international community should aspire to. A consensus has not yet formed around the desirability of complete nuclear disarmament. In an important respect, this should affect neither the shaping of policies on the regulation of fissile materials, nor the energy devoted to that task. This is because the quantities of fissile materials outside nuclear weapons are now so great, in absolute terms and as a proportion of total inventories, that today's need for the stringent control and regulation of nuclear materials and facilities is little different from that which would exist in conditions of total disarmament. As a consequence, the regulatory situation in all countries should be approached as if the world is preparing for nuclear disarmament, whether or not that is in fact the case.

In another respect the confusion does matter, and for reasons that go well beyond the current linkage of the FMCT negotiations to the disarmament issue. To what degree should the retention of nuclear arms determine the regulatory stance adopted today and in the future by nuclear weapon states? Can the tail be allowed to wag the dog? These questions should lie, in my view, at the centre of our concerns. If the de jure and de facto nuclear weapon states are not yet prepared to embrace nuclear disarmament, then it is incumbent upon them to develop a much clearer conception of how fissile materials in a partially disarmed state should be regulated. If this issue is not addressed, we shall be forced to live with constant uncertainty, lethargy and dissatisfaction - and with an undeniable risk of reversal.

The second point concerns the need for policy drivers and treaty frameworks. Many steps, large and small, have been taken in recent years that help to increase irreversibility. Their value should not be underestimated. But beyond the 93+2 Programme, which was directed mainly at non-nuclear weapon states, no multilateral treaty negotiations have been initiated on fissile materials. ₈ The current obstruction of the FMCT negotiations is regrettable both because it prevents moves towards adoption of a new international norm (the production ban), and because it robs us of an essential policy driver. Although the FMCT's scope and objectives may be constrained, all parties in the negotiations will find themselves having to address fundamental regulatory issues. The FMCT will help drive the re-regulation of partially disarmed states that I have alluded to above.

However, the FMCT cannot achieve this on its own. There needs to be a complementary process attending to stocks (it is possible that START III might form part of that process). To anyone that has studied those stocks, it is inconceivable that they should be shunted into political and regulatory sidings. In support of these policy initiatives, there is also need for a more focused discussion of principles. Five principles have been tentatively proposed above, under the headings irreversibility, transparency, minimization, equality of verification standards, and permanence of verification. Whether or not these are considered appropriate, it is important to recognize that principles need to be enunciated if there is to be clarity of purpose.

Finally, it is worth drawing attention again to the 'predisposition to opaqueness' that characterises processes of armament, and the 'predisposition to transparency' that characterises processes of disarmament. In countries with nuclear weapon programmes, there is now uncertainty and disagreement, both within governments and between governments, over where to draw the boundaries between opaqueness and transparency. If the irreversibility of arms reductions and disarmament measures and processes is to be deeply embedded, a general 'presumption to disclose' will have to replace the 'presumption to withhold' that has prevailed to date.

Notes:

Note 1: Judgements are usually relative. The 'superior condition' may still be far from ideal, but a great improvement upon the inferior condition against which it is measured. In many contexts (e.g. South Africa and the Ukraine today), the awfulness of prior conditions may commit people to the irreversibility of imperfect current conditions. Back.

Note 2: Changes in values come about, of course, partly through the successful development of regimes. The non-proliferation regime has, for instance, greatly widened attachment to 'non-nuclear cultures'. A fine recent analysis of regime development and compliance can be found in Abram and Antonia Handler Chayes, The New Sovereignty: Compliance with International Regulatory Agreements, Harvard University Press, 1995. Back.

Note 3: 'Evolutionary versus planned approaches to nuclear disarmament', Disarmament Diplomacy, May 1997. For these reasons, the current advocacy of time-bound disarmament by India and some other non-aligned nations, and their insistence that there should be universal endorsement of it before they will allow the FMCT and other measures to be developed, are in my view both misguided and counter-productive. Back.

Note 4: There were, however, a few exceptions. Under the Hexapartite Safeguards Project, six countries (Australia, Germany, Japan, Netherlands, UK and USA) committed themselves to placing centrifuge enrichment plants on their territories under permanent IAEA safeguards. Japanese fuels sent to Britain and France for reprocessing also have to be kept permanently under IAEA safeguards in compliance with INFCIRC/66-type agreements that remain in force. Back.

Note 5: In China, however, all (or nearly all) inventories of civil plutonium are, to my knowledge, being safeguarded by the IAEA today. The main reason is that China has agreed with foreign reactor suppliers, and with foreign suppliers of components to indigenous reactors, that plutonium produced in the reactors will be submitted to safeguards. The new centrifuge enrichment plant being supplied by Russia will also be safeguarded by the IAEA. The safeguards situation in China is therefore beginning to resemble that in France and the UK. Back.

Note 6: See D. Albright, F. Berkhout and W. Walker, Plutonium and Highly Enriched Uranium 1996, Chapter 15, pp. 433-437. Back.

Note 7: On the weapon register, see Harald Müller, 'Transparency in Nuclear Arms: Towards a Nuclear Weapons Register', Arms Control Today, October 1994. Back.

Note 8: The International Management discussions in Vienna have involved negotiations between officials appointed by nine governments, but the outcome will not take the form of a treaty and will involve guidelines rather than obligations. See P. Agrell, 'Guidelines for the responsible management of plutonium', paper presented to the International Symposium on Nuclear Fuel Cycle and Reactor Strategies, Vienna, 3-6 June 1997, IAEA-SM-346/114. Back.