Five Plus Three: How to Have a Meaningful and Helpful Fissile Material Cutoff Treaty

The Obama administration has stated its intention to conclude a treaty cutting off production of fissile material, highly enriched uranium (HEU) and plutonium, for nuclear weapons. So did the administrations of George W. Bush and Bill Clinton. Although a fissile material cutoff treaty (FMCT) has been a key objective of the UN Conference on Disarmament (CD) for many years, that organization seems unable to break out of its now customary paralysis.[1] This experience should encourage the Obama administration, if indeed it wants such a treaty, to look to another forum for realization. More importantly, it may be vital for the new U.S. administration to take a different approach because pursuit of an FMCT as currently contemplated at the CD might well have the ironic and presumably unintended consequence of gravely undermining the nuclear nonproliferation regime. Fortunately, such a new approach may well be available.

The Current FMCT Impasse

The current impasse at the consensus-based CD began a decade ago. A significant part of the blame for the collapse of initial negotiations on an FMCT should be apportioned to the Nonaligned Movement (NAM)-never a group to pass up a chance to impede something good by insisting on what it thinks would be ideal-for trying to force greater attention on accelerating disarmament. China and Russia are similarly at fault for preventing progress on such talks by trying to link FMCT discussions to commencing negotiations on their disingenuous draft of a treaty banning space weapons. That initial Chinese-Russian linkage of unrelated issues grew into the laundry list of pet arms control projects known as the A-5 proposal containing, amazingly, negotiating mandates for four major treaty efforts (disarmament, negative security assurances, space weapons, and an FMCT), which most CD members now recognize as completely unworkable as a practical matter.

After the collapse of the A-5 proposal under its own weight, the formula advanced in the document L.1, now embodied in document CD/1840, became the current focus of debate at the CD. This approach still supports "discussion" mandates for the other issues but more feasibly proposes a negotiating mandate only for an FMCT. Even progress on this more modest document, however, has so far been impossible. Specifically, China and Pakistan-the former being the only nuclear Nonproliferation Treaty (NPT) nuclear-weapon state to be increasing its nuclear arsenal rather than reducing, the latter apparently still too concerned about India to be very interested in restraint-have blocked negotiations on a treaty. Perhaps not surprisingly, they have been joined by Iran in their opposition to negotiating an FMCT. So far, none of these three countries is apparently interested in foregoing the option of producing fissile material for nuclear weapons.[2]

Although only Beijing, Islamabad, and Tehran remain as obvious obstacles to the commencement of FMCT negotiations, with Jerusalem and New Delhi remaining more circumspect with regard to their intentions, many hurdles still lie ahead with regard to their conclusion. Among these challenges is the issue of verification. After 1995, the CD's agreed negotiating parameters for an FMCT were defined by the so-called Shannon Mandate, which required pursuit of a treaty that was "non-discriminatory, multilateral and internationally and effectively verifiable."[3] This goal was accepted by the Clinton administration and may yet be supported by the Obama administration.[4] The Bush administration concluded in 2004 that no FMCT, regardless of how costly and intrusive a verification system one postulates, much less an FMCT likely to be agreed on by all parties at the CD, would likely be effectively verifiable. In Washington's view, this should not have presented an obstacle to commencing FMCT negotiations and indeed concluding a normative FMCT, but it did require states to agree to revising the Shannon Mandate's requirement to achieve such verifiability.

The Shannon Mandate required, up front, that any concluded FMCT be both universal and nondiscriminatory-that it not apply different rules to different countries, which was a continuing source of criticism about the NPT-and effectively verifiable. Although the Bush administration was not averse to discussing verifiability during FMCT negotiations and U.S. officials did not require ruling out ever adopting some verification measures, Washington deemed it improper for the mandate to require the achievement of what was arguably unachievable as a precondition for even beginning negotiations. Today, except for China and Pakistan, all parties at the CD apparently stand ready to begin negotiations that all assume will include extensive discussions of and efforts to achieve at least some degree of verifiability.

The issue of verifiability will not go away, therefore, and it is likely that even if negotiations are started on an FMCT, many disputes and problems will arise over whether effective verifiability can be achieved and, if so, whether the cost of doing so is one that all CD negotiators will prove willing to pay.

Technical Issues Associated With Verification

Today, U.S. government verification experts have apparently made enough headway in explaining their position that many CD delegations have conceded privately that effective verification is indeed impossible. This has not dimmed the enthusiasm of many of them for adopting some verification measures and for deeming even partial measures to be "good enough." (In this author's presence, in fact, some have actually advocated pretending that whatever measures end up being adopted in fact amount to effective verification, although no one seems to have been so bold as to say this publicly.) To the extent that effective verification remains an important objective of negotiations, one can expect that the road ahead for an FMCT will remain a rocky one.

The United States has explained its verification arguments on many occasions, in print in the context of NPT review-cycle debates[5] and in person through numerous expert-level and diplomatic briefings and discussions with CD partners. It is beyond the scope of this article to detail all the arguments, but some of the salient points from a technical perspective include the difficulty for the verifier of (1) conducting adequate verification inquiries at facilities associated with nuclear weapons work or other sensitive activities without compromising the host government's national security information, (2) dating the point of first "production" for plutonium that has been reconditioned subsequent to its initial separation, (3) ascertaining the true intention behind fissile material production under a treaty that will not ban all such production but rather only that done for purposes of use in nuclear explosives, and (4) telling anything useful from environmental sampling at nuclear weapons facilities where fissile material has been produced for years and particulate contamination is already ubiquitous. Conceivably, there could be clever ways around these problems, but such solutions apparently have yet to be offered, and effective verification seems as distant a dream as ever.

Policy and Programmatic Issues for Verification

An FMCT's problems, however, are only partly technical. Some of the difficulties are more political in nature. An FMCT would clearly be more verifiable, which is not the same thing necessarily as thinking it effectively so, if all existing stocks of fissile material were to be subject to examination and monitoring under an FMCT verification regime. Because such stocks include material actually used in nuclear weapons by possessor states, kept as raw material within their weapons programs, or held for highly sensitive purposes not related to nuclear explosives, such as fuel for naval nuclear reactors, however, it is very difficult to imagine any FMCT could be concluded by consensus that actually covered all existing stocks in a meaningful way. (One must also guard against having an FMCT verification protocol become a mechanism for proliferating nuclear weapons know-how.)

Much of this problem would go away were it possible to apply different rules to states possessing nuclear weapons than to nonpossessors, as does the NPT, but this would fly in the face of the widespread desire that an FMCT be nondiscriminatory. In other words, within the constraints of the political requirement that uniform rules must apply to all parties, an FMCT could be either comparatively verifiable by virtue of covering all existing stocks or it could be negotiable, but probably not both. (The possessors of nuclear weapons presumably will not all be willing to permit intrusive inspections of their weapons programs for so long as such arsenals remain. This is not a law of nature, of course, and some in the arms control community apparently hold out high hopes that possessors will suddenly decide that it is no longer important to conceal such details from each other and from nonpossessors. My assumption, however, is likely to hold true for a long time.)

Yet, the problems do not stop there. From my discussions with participants in recent FMCT debates, it has become clear that further problems lie ahead for the current CD approach to a fissile material treaty, including the danger that it could undermine the NPT itself. This issue has not so far been the subject of much if any public discussion. Nevertheless, such worries underlie some countries' grave concerns about any effort to achieve FMCT verification within the context of the treaty's still presumptively universal application and likely nondiscriminatory character.

An FMCT verification regime, it is feared, might undermine the NPT regime in three ways.

1. It might tend to legitimate the development of "peaceful" fissile material production capabilities in ways incompatible with the survival of the nonproliferation regime.

2. An FMCT verification system, if pursued on the basis of universality and nondiscrimination, might seem to offer an alternative but weaker standard of verification that would undercut and in practice replace adherence to the 1997 Model Additional Protocol of the International Atomic Energy Agency (IAEA), particularly in worrisome states such as Iran.

3. Debates over FMCT verification with regard to existing fissile material stocks and fissile material held for sensitive but nonexplosive purposes, such as fueling naval reactors, would draw unhelpful attention to a long-standing potential loophole within the IAEA-run verification system for Article III safeguards under the NPT. Agreement on an FMCT verification regime could worsen this problem and generalize this loophole by the de facto adoption of a standard that would not merely leave the FMCT unverifiable but would also undercut existing approaches to nuclear safeguards.

Legitimating Ubiquitous Fissile Production

The first concern is more symbolic and purely political and relates to ongoing disputes over the wisdom of permitting, not to say facilitating, the further expansion of fuel cycle capabilities around the world. It has long been understood that fissile material availability is the principal hurdle to be overcome for any country seeking to develop nuclear weapons and the major pacing element for a weapons program. This is why, for instance, U.S. assessments of Iran's reported suspension in 2003 of certain weaponization-related aspects of its nuclear weapons effort in response to international pressure[6] have done so little to allay concerns about Tehran's nuclear weapons ambitions. As Director of National Intelligence Michael McConnell put it, Iran seems to have suspended only what is "probably the least significant part of the program."[7] Once fissile material production capability has been obtained, the rest is comparatively easy, quick, and concealable. Iran's uranium-enrichment program has underlined these concerns and spurred the Bush administration to place even greater emphasis on preventing the spread of uranium enrichment and spent fuel reprocessing (for plutonium) to additional countries. By threatening to create a world in which many countries possess a rapid and easy nuclear option, the spread of enrichment and reprocessing technology presents a grave challenge to the nonproliferation regime.

In this context, the advent of an FMCT would not necessarily be an unequivocal good. As it is currently contemplated, an FMCT would not ban the production of all fissile materials. Instead, current debates at the CD revolve around a draft FMCT, offered by the United States, that would prohibit only "produc[ing] fissile material for use in nuclear weapons or other nuclear explosive devices, or us[ing] any fissile material produced thereafter in nuclear weapons or other nuclear explosive devices."[8]

Some have expressed concern that achievement of an FMCT, by seeming to ensure that fissile material will not thereafter be produced for explosive purposes, would be perceived to guarantee the permissibility and the harmlessness of fissile material production for other purposes by essentially anyone else. After all, the NPT does not speak directly to the issue of fissile material production. Even Article IV, which is often taken, in my view mistakenly, to convey or acknowledge some kind of general enrichment and reprocessing rights, does not directly address the question. The broad assessment of Article IV issues offered by the Bush administration in 2005 therefore remains correct:

Some have asserted that any State Party in demonstrable compliance with the NPT has a specific right to develop the full nuclear fuel cycle, and that efforts to restrict access to the relevant technologies is inconsistent with the NPT. The Treaty is silent on the issue of whether compliant states have the right to develop the full nuclear fuel cycle, but...it does provide for discretion on the part of supplier states regarding the nature of their cooperation with other states.... While compliant State[s] Party should be able to avail themselves of the benefits that the peaceful use of nuclear energy has brought to mankind, the Treaty establishes no right to receive any particular nuclear technology from other States Party-and most especially, no right to receive technologies that pose a significant proliferation risk.[9]

An FMCT, however, might reshape the still-contested terrain of contemporary enrichment and reprocessing debates by more directly addressing the issue of fissile material production. If a formal treaty had been concluded to prohibit fissile material production for explosive purposes, it might seem obvious that there existed a right to produce fissile materials for nonexplosive purposes and all that much less reason to object should any FMCT state party propose to do so. As a consequence, an FMCT could contribute to the proliferation of sensitive capabilities that would result in more countries attaining the option of acquiring nuclear weapons virtually at will. To say the least, it is not obvious that an FMCT that notionally banned fissile material production for weapons purposes but which in fact contributed to the global ubiquity of an enrichment and reprocessing-facilitated nuclear weapons option would be a net benefit for global security.

Undermining the Model Additional Protocol

The second concern relates to the danger that an FMCT verification regime would undermine efforts to achieve broader adherence to and ultimately the universality of the IAEA Model Additional Protocol. The model protocol was drafted in the wake of revelations after the 1991 Persian Gulf War about how close Iraq had come to developing nuclear weapons under the very noses of IAEA inspectors. Those inspectors had relied on monitoring declared nuclear material, facilities, and activities conducted pursuant to traditional comprehensive safeguards agreements reached with the agency. These safeguards, however, were wholly inadequate to the challenge of detecting undeclared activities in any country that might wish secretly to use nuclear technology for illicit purposes. The subsequent protocol expanded the reporting requirements of member states and gave inspectors additional access rights in order to determine the absence of undeclared activities.[10] A critical component of the new protocol was the right of agency inspectors to collect environmental samples.

To date, adherence to additional protocols is still far from universal, having been entered into force by only 89 states. To be sure, the IAEA director-general and its board of governors have made clear that even an additional protocol's new authorities are inadequate to the challenges presented by IAEA verification activities in the face of sophisticated denial and deception activities undertaken by countries such as Iran.[11] Nevertheless, if nuclear safeguards are to provide any sort of meaningful barrier against nuclear weapons development, an additional protocol is widely and correctly regarded as being essential to the survival of the safeguards system.

There is concern, however, that an FMCT verification regime could undermine the prospects for universalizing the Model Additional Protocol by appearing to provide an alternative and most assuredly weaker international nuclear safeguards baseline to which countries might tend to adhere. The protocol is aimed at allowing the IAEA some chance to ferret out undeclared activities in non-nuclear-weapon states; it does not have to worry about applicability to nuclear weapons possessors with a legitimate need to protect nuclear weapons-related national security equities and restrict the dissemination of proliferation-sensitive information.[12]

Moreover, the authorities an additional protocol provides to inspectors are reasonably broad and intrusive and impose no small burden, in political terms and with regard to financial and administrative costs, on countries adhering to it. Given the costs, some countries, especially in the developing world, have resisted agreeing to an additional protocol and resent suggestions that it be made a precondition for the provision of nuclear technology pursuant to Nuclear Suppliers Group rules.[13] (Other countries, most obviously Iran, have less legitimate reasons for resisting and resenting the Model Additional Protocol's focus on trying to detect undeclared nuclear activities, but oppose it they do.) Universalizing the protocol is strongly supported by many supplier states and the IAEA as a measure indispensable to preserving the integrity of the nonproliferation regime, but it remains somewhat controversial.

Because an FMCT, as currently contemplated in the CD, is expected to be universal and nondiscriminatory, its verification mechanisms, as noted above, would of necessity be weaker than those of the Model Additional Protocol. It might be possible to set up a two-tier FMCT verification system that would treat possessors of nuclear weapons differently than all others in the international system, which is done for IAEA safeguards under NPT Article III, including in connection with the Model Additional Protocol; but in practice this would be politically difficult, not to mention incompatible with the idea of nondiscrimination as it is usually employed.[14] More likely, FMCT verification rules would have to be developed on a lowest-common-denominator basis through the consensus-based CD, with the result that inspector access and other authorities established for purposes of FMCT verification would probably end up being a pale reflection of those provided in the Model Additional Protocol.

If an FMCT offered a verification program less burdensome and effective than Model Additional Protocol safeguards, states might find adherence to an FMCT a tempting excuse to decline or repudiate the protocol. The fact that such a country had signed on to FMCT verification provisions also would make it correspondingly more difficult, as a political matter, to pressure that country to agree to the Model Additional Protocol. In a world of finite financial resources and political capital, it might be difficult to maintain two parallel and independent systems of international verification, particularly if the IAEA were itself to be entrusted with most of the work of FMCT verification, as is frequently discussed in CD circles.

Some participants in FMCT debates thus worry that a more lenient FMCT verification system might come to undermine and eventually replace today's emerging, Model Additional Protocol-based system of nuclear safeguards. This outcome would not necessarily take too much time to manifest itself either. The countries about which one should worry the most and whose adherence to the protocol would be most valuable-Iran comes to mind once more-would presumably also be the ones most likely to jump ship quickly in favor of a weaker FMCT verification regime. The replacement of the Model Additional Protocol approach by a less stringent FMCT system would therefore probably undermine nonproliferation efforts and would poorly serve the interests of international peace and security. The Hippocratic oath is perhaps here instructive: "First, do no harm."

Cementing in Place a Safeguards Loophole

A third concern with an FMCT verification system relates to its interaction with a potential loophole in current comprehensive safeguards agreements required by NPT Article III and currently in place for all but 30 NPT states party. The IAEA information circular INFCIRC/153 provides the model for all such agreements presently in force. Paragraph 14 of the first part of INFCIRC/153 provides for the possibility that a country subject to a comprehensive safeguards agreement will "exercise its discretion to use nuclear material which is required to be safeguarded thereunder" for a "non-proscribed military activity."[15] This model agreement thus theoretically allows countries to remove nuclear material from safeguards for purposes of using it for a nonpeaceful purpose that is nonetheless not prohibited by the NPT. This could presumably cover a range of potential uses for nuclear material, provided that these uses do not relate to nuclear explosives, but Paragraph 14 is most obviously applicable to the use of enriched uranium in the nuclear reactors used to power naval vessels.

Provided that this provision is indeed only used in connection with legitimately sensitive activities such as naval nuclear propulsion and only by states already possessing nuclear weapons, it need present no fundamental problem for the IAEA's system for nuclear safeguards verification. The provision is potentially subject to enormous abuse, however, because a rush of NPT non-nuclear-weapon states suddenly declaring or pretending a sudden interest in naval reactor development would result in the removal of nuclear material from safeguards in quantities more than sufficient to conceal nuclear weapons work prohibited by NPT Article II. Abuse of Paragraph 14 could therefore create a vast hole in the safeguards system.

The specific parameters and scope of acceptable uses for Paragraph 14 have not been defined, but its provisions have not emerged as a problem for safeguards verification largely because Paragraph 14 issues have not been the focus of much attention or debate.[16] Efforts to develop a universal and nondiscriminatory FMCT verification protocol could change this. It seems unlikely that countries engaged in sensitive but nonproscribed military activities such as naval nuclear propulsion would all be willing to open such programs to detailed examination by FMCT verifiers. (If these activities were not militarily sensitive, the issue would not arise in the first place.) Consequently, an FMCT verification protocol would probably have to have some kind of verification exemption for such activity, just as it would presumably also contain an exemption for nuclear weapons-related activity in the nuclear-weapon states. (The necessity of creating such exclusions, which would presumably be a precondition for acceptance of the treaty by countries engaged in such work, is another reason why it is so difficult to imagine an FMCT being genuinely verifiable.) In the course of debating and codifying universal and nondiscriminatory rules for FMCT verification, attention would likely be drawn to the issues of verification-exempt sensitive but nonproscribed activities and a set of understandings reached about their scope and nature as things in which any country may engage. Otherwise, an FMCT verification protocol would be discriminatory.

There is, therefore, reason to fear that FMCT verification efforts could damage IAEA nuclear safeguards verification under the NPT, by effectively flagging Paragraph14 as a safeguards loophole and providing the international community with a de facto working definition of its parameters that could gravely undermine the effectiveness of verification work pursuant to comprehensive safeguards by allowing essentially any country to remove nuclear materials from safeguards on the pretense of working on naval nuclear reactors or other such programs. FMCT verification negotiations could become the vehicle through which a provision unproblematically used only by nuclear-weapon states to permit the continuation of legitimate activities metastasizes into an NPT-destructive means by which non-nuclear-weapon states could easily defeat IAEA safeguards verification and conceal prohibited nuclear weapons work.

An Alternative Route to a Meaningful FMCT?

What is one to do if one still hopes for some kind of a ban on the further production of fissile material for nuclear weapons? Must we give up on the idea of an FCMT? The search for a way to surmount or sidestep these many challenges might begin with an admission that a treaty banning the production of fissile material for nuclear explosive purposes would add nothing of significance to the nonproliferation obligations of any NPT state party that does not already lawfully possess nuclear weapons.

After all, NPT non-nuclear-weapon states are already prohibited by NPT Article II from "manufactur[ing] or otherwise acquir[ing] nuclear weapons or other nuclear explosive devices,"[17] a rule that has long been understood to cover not merely the end stage of "manufacturing" a nuclear device but also such prior steps as may have been undertaken in service of this illicit purpose.[18] Producing fissile material for a nuclear weapon unquestionably counts as a violation under this standard. The only reason that an FMCT would be interesting or valuable, therefore, lies in its potential to constrain fissile material production for weapons purposes in the NPT nuclear-weapon states[19] and in countries such as India, Israel, and Pakistan presently outside the NPT but which have demonstrated or are presumed to have nuclear weapons. For these states, an FMCT would be a significant step because it would formally constrain the size of their nuclear weapons programs for the first time.

As the foregoing discussions indicate, however, attempting to craft a universal and nondiscriminatory FMCT verification regime would run into a great many problems in trying to fit one-size-fits-all provisions to the very different situations presented by the presence in the international community of weapons-possessing and non-weapons-possessing countries. What if one were to relax the reflexive requirement enshrined in the Shannon Mandate and retained in the negotiating mandate of CD/1840 that an FMCT be nondiscriminatory and universal?

As suggested earlier, one way to approach an FMCT more seriously from a verification perspective might be to permit it to become discriminatory like the NPT: treating possessors of nuclear weapons differently than nonpossessors. An alternative approach, however, and conceivably both a more efficient and ultimately a better one, might be to jettison the ambition to universality. Because an FMCT would not add meaningfully to the obligations of NPT non-nuclear-weapon states anyway, why bother with all the trouble, uncertainty, and negotiating headaches of developing an FMCT in a consensus-based forum such as the 65-member CD when all one really wants to do is reach eight states (China, France, India, Israel, Pakistan, Russia, the United Kingdom, and the United States)?[20]

Limiting one's FMCT approach to the "Five Plus Three" countries-the five NPT-recognized nuclear-weapon states plus the three outliers who have not adhered to the NPT-would have an admirable parsimony. It would tailor the legal remedy precisely to the wrong of fissile material production for nuclear explosive purposes in the only countries for which this is not already prohibited by international law. It would do so without having to wrestle with accommodating an additional four or five dozen members of the CD in consensus-bound multiparty negotiations or to sweep into a treaty some 180 NPT non-nuclear-weapon states to whose legal obligations an FMCT would add in no meaningful way, but whose IAEA safeguards obligations might thereby effectively be weakened.

Restricting an FMCT to the Five Plus Three countries might actually help the treaty sidestep some of the formidable verification hurdles it would otherwise face in a world that contains countries as vastly differently situated, in nuclear weapons terms, as the United States and Swaziland. It is not that any fewer accommodations detrimental to verification confidence would be necessary to make an FMCT acceptable to countries engaged in nuclear weapons development and often in sensitive nonproscribed military activities. What would distinguish a Five Plus Three treaty from a universal FMCT is the degree to which the uncertainty created by such exemptions would matter.

Effective verification is not a hard science or a subject well suited to quantification and rigid, bright-line rules. Levels of verification uncertainty unacceptable in one context may be quite tolerable in another. To contrive a simplistic example, the inability of verifiers to rule out a half-dozen nuclear warheads remaining hidden in a country subjected to arms control limits would have dramatically different consequences if the limits in question were at the level of 1,000 warheads or at the abolition level of "zero." A hypothetical six-warhead margin of error might be acceptable in the former case, perhaps amounting to "effective verification," but could be a catastrophic level of uncertainty in the latter. (Retaining a handful of concealed warheads in a disarmed state or acquiring them in any state under an abolition regime would be a strategic coup of great magnitude.) The key in verification analysis is therefore sometimes not the specific error margin per se, but policy judgment calls weighing such factors as the likely impact of a violation on the military balance being regulated by the legal norm,[21] the ease and speed with which other parties could respond to or remedy a violation after its detection, or the degree to which parties have incentives to cheat in the first place.[22]

It might also be the case that confining an FMCT to states already possessing a demonstrated or presumed nuclear weapons capability might at least somewhat lessen the daunting problems presented by verification uncertainty. The likely necessity of having verification exemptions for sensitive nonproscribed activities, for instance, would gravely undermine a universal FMCT by making it vastly more difficult to ascertain whether prohibited activity-nuclear weapons work-was in fact occurring in a non-nuclear-weapon-state. In a regime confined to the eight presently weapons-capable countries, however, it might be slightly less disastrous from a nonproliferation perspective to bow to international pressure to settle for some verification or verification that is "good enough," inasmuch as uncertainty with regard to countries that already have nuclear weapons would not open the door to secret weapons work in non-nuclear-weapon states. At least at nuclear-weapon-state warhead levels not drastically lower than those of today, at any rate, a degree of verification uncertainty in an FMCT for the Five Plus Three states would imperil global security less than it would under a universal and nondiscriminatory FMCT.[23]

Once one escapes from the theological presupposition that an FMCT must be universal and nondiscriminatory, in fact, there may be additional advantages in being willing to employ such flexibility, in pursuit of a meaningful and workable treaty even among the Five Plus Three countries. In other words, one verification size need not necessarily fit all. For example, in the case of Russia and the United States, large quantities of fissile material are already being removed from nuclear weapons-related stockpiles and converted to peaceful uses. For these two states, no verification procedures would really be needed-at least for a long time-in order for the international community to have confidence that they were not producing additional fissile material for use in nuclear weapons: the continuing flow of material out of weapons-related stocks into peaceful applications would itself surely be evidence enough that Moscow and Washington felt no need for more fissile material for weapons. In the case of France, the United Kingdom, and the United States, very comprehensive voluntary safeguards agreements are also in place that cover nuclear material in peaceful programs. With respect to India and Pakistan, an arrangement similar to that between Argentina and Brazil might help resolve any concerns that these two countries have with respect to each other. These examples do not resolve all possible concerns, of course, but they illustrate how pragmatic arrangements could perhaps be made that would make an FMCT meaningful without trying to find a single solution that would apply to all parties. Even in the context of a "Five Plus Three" treaty, trying to achieve a general solution could lead to extended and unnecessary negotiations.

Why then should one not pursue an eight-party FMCT? At least two objections might be offered. First, it might be contended that it would be inappropriate for the FMCT issue to be taken away from the CD in Geneva. Second, it might be argued that it would be inappropriate to recognize the "non-NPT Three" (India, Israel, and Pakistan) as having a sort of quasi-nuclear-weapon-state status by engaging specifically with them in FMCT negotiations and not approaching the question on a universal basis.

The first of these concerns may most easily be dismissed. To suggest that an FMCT be made universal so that diplomats at the CD can have something with which to occupy themselves is, to put it charitably, putting the cart before the horse. The CD should pursue specific arms control and disarmament proposals when and where its involvement is necessary or helpful for progress on agreements that serve the interests of international peace and security. If the question at hand is to negotiate a universal agreement, such as the Chemical Weapons Convention, a past CD success story, then the conference is presumably at least as good as any other universal negotiating forum available and no doubt rather better than the unfocused and erratic UN General Assembly or UN First Committee.

As we have seen, however, there is no functional necessity for an FMCT to be universal, but there is good reason to approach such a treaty directly and efficiently by addressing it only to the eight countries whose participation would matter in the slightest. If it is desirable to secure full employment for diplomats in the comforts of Geneva irrespective of their ability to reach agreement in the long-paralyzed CD, there are no doubt other universal proposals that could be discussed with satisfying unproductiveness. If one wishes actually to achieve an FMCT, it might be advisable to take discussion directly to the relevant players in a context that stands a reasonable chance of producing results. That said, it was precisely over such concerns of being attacked for killing the CD that the Bush administration chose not to pursue this most logical of options. U.S. officials did not wish to give ammunition to those who accused the administration of being hostile to multilateral approaches and institutions. It may be that such uncharacteristic squeamishness was a mistake and that a Five Plus Three approach outside the CD represents a better route to an FMCT and to improved global security.

The objection focusing on the danger of recognizing India, Israel, and Pakistan as quasi-nuclear-weapon states is a more creditable one. Conducting FMCT negotiations on a Five Plus Three basis is admittedly at least somewhat like according these three countries some special status on account of their actual or presumed possession of nuclear weapons. Nevertheless, addressing an FMCT to the only countries for which it is needed-namely, the eight countries that are not NPT non-nuclear-weapon states-is not the same thing as recognizing the three NPT nonmembers among them as NPT nuclear-weapon states. Moreover, one need not necessarily enter into Five Plus Three FMCT negotiations on an eight-party basis. Indeed, the best approach may be to build on what we already have. The draft FMCT tabled by the United States in the CD already contains almost all of the features needed for a meaningful treaty, especially if it were limited to the eight states. Leaving aside any debate that might take place over verification measures, which could in any case be handled in an FMCT protocol or in one or more other side agreements rather than in the treaty itself, the only modifications that might be needed to the current U.S. draft would be with respect to the entry into force and duration provisions.

The duration of a Five Plus Three FMCT, as it applies to any state party, should be such that the treaty would cease to apply to a state that joins the NPT as a non-nuclear-weapon state. This duration provision would thus make clear that an FMCT is not an end point but rather a step toward the global applicability of the NPT and perhaps indeed an eventual end to the existence of nuclear weapons. If additional verification measures are felt necessary beyond the provisions currently in the U.S. text, these could be negotiated separately among the states that felt that such additional measures were needed. For example, the five NPT nuclear-weapon states may wish to reach an agreement as to what should be covered in their voluntary safeguards agreements. As mentioned earlier, India and Pakistan might feel that they needed additional arrangements to address their concerns. This approach would let verification provisions be tailored to the verification needs of each party, if any, without having to work out a one-size-fits-all approach that applies uniformly and coherently to possessors of nuclear weapons as vastly differently situated as the various members of the eight.

I do not mean to suggest that this approach is a panacea. If China and Pakistan continue to wish to preserve for themselves the ability to produce additional fissile material for nuclear weapons, for instance, they will presumably continue to try to block progress on any sort of FMCT in the CD or elsewhere. After all, diplomats' instinctive faith in the power of diplomacy to improve the world should not be allowed to become naivete: where interests fundamentally diverge, fine talk and clever drafting cannot be expected to bridge them. Pursuing a flexible approach outside the CD, however, may offer the best chance there is for an FMCT. If there are indispensable parties who remain unalterably opposed, there will be no treaty. Nevertheless, even then there might be some value in flushing out objectors, forcing them to reject a sensible and accommodating substantive proposal in the open, rather than sheltering behind procedural gamesmanship and thinly rationalized issue linkages in a consensus-based institution such as the CD. There are worse things than clarity.

Conclusion

The CD is today closer to the commencement of negotiations on an FMCT than at any time since the original negotiations collapsed more than a decade ago. Closer does not necessarily mean close, however, and enormous challenges remain to be overcome before any FMCT would likely be able to emerge. Significantly, these challenges include not merely difficulties related to verifiability but also grave and publicly ignored policy and programmatic issues related to the potential impact of FMCT verification on the nuclear safeguards system. An FMCT could dangerously undermine the existing verification system for NPT Article III safeguards and the IAEA Model Additional Protocol.

It might be politically controversial and some would surely allege that it amounted to a vote of no confidence in the CD; however, an approach to the FMCT question that focused only on the five NPT nuclear-weapon states and the three non-NPT outliers may offer the best chance for progress in light of these problems and for an FMCT that, in the end, genuinely contributes to global security. Arguably far more than current CD negotiations, a Five Plus Three approach offers the chance to achieve an FMCT that is realistically negotiable and meaningful. Can anyone genuinely interested in an FMCT afford to ignore this potential solution?

Click here to comment on this article.

Christopher A. Ford is a senior fellow and director of the Center for Technology and Global Security at the Hudson Institute in Washington, D.C. Ford served as U.S. special representative for nuclear nonproliferation and as a principal deputy assistant secretary of state for verification, compliance, and implementation during the George W. Bush administration.

ENDNOTES

1. For a window on recent debates over the future of the CD, see Stephen G. Rademaker, "The Conference on Disarmament: Time Is Running Out," Arms Control Today, December 2006, pp. 13-15; Paul Meyer, "The Conference on Disarmament: Getting Back to Business," Arms Control Today, December 2006, pp. 16-17; Michael Krepon "The Conference on Disarmament: Means of Rejuvenation," Arms Control Today, December 2006, pp. 18-22.

2. When pressed on the issue of fissile material production, Chinese officials, who seem to hate being seen as an obstacle to arms control progress, even or perhaps especially when they are, sometimes privately claim that China has ceased production. It is not clear whether this is true, although it conceivably could be if Beijing has already produced and stockpiled enough fissile material to support whatever plans it may have for its still expanding arsenal. On the other hand, some Chinese have also privately assured me that they are also not actually expanding their arsenal, which seems to be entirely untrue. At any rate, China has thus far refused to allow itself to be held publicly accountable for the claim.

3. "Report of Ambassador Gerald E. Shannon of Canada on Consultations on the Most Appropriate Arrangement to Negotiate a Treaty Banning the Production of Fissile Material for Nuclear Weapons or Other Nuclear Explosive Devices," CD/1299, March 24, 1995.

4. One former Clinton administration official recently urged the United States to reverse its position on FMCT verification, although he mischaracterized Bush administration policy as per se "oppos[ition to] international monitoring" rather than more accurately describing it as opposing the Shannon Mandate's presupposition of effective verifiability. Robert J. Einhorn, "Controlling Fissile Materials and Ending Nuclear Testing," Paper presented to the "Achieving the Vision of a World Free of Nuclear Weapons" conference, Oslo, February 26-27, 2008, p.3. In fact, U.S. officials at the CD under Bush opposed neither discussion of verification issues nor necessarily even the eventual adoption of some verification measures. They merely felt that effective verification of an FMCT was unachievable and refused to agree to a mandate requiring it. Einhorn believes that in general "some monitoring tasks" under an FMCT are "no more difficult than monitoring compliance with the NPT" and that the IAEA could "do the job effectively." Einhorn's qualifier "some" is suggestive, however; and his account provides no explanation of how other FMCT tasks, such as ascertaining the purpose for which undeclared fissile material had been produced, might be effectively accomplished. See Christopher Ford, "The United States and the Fissile Material Cutoff Treaty," Paper presented to the "Preparing for 2010: Getting the Process Right" conference, Annecy, France, March 17, 2007, www.state.gov/t/isn/rls/other/81950.htm (noting problem of ascertaining purpose of material production, a challenge that is "qualitatively different from those involved in verifying compliance with IAEA safeguards"). At any rate, citing the continued unwillingness of China, India, and Pakistan to forswear production of fissile material for weapons purposes, Einhorn contended that "entry into force of an FMCT is several years away at a minimum" and apparently worries that an FMCT could "remain[] deadlocked." Einhorn, "Controlling Fissile Materials and Ending Nuclear Testing," pp. 4, 6.

5. See Ford, "United States and the Fissile Material Cutoff Treaty."

6. See U.S. National Intelligence Council, "Iran: Nuclear Intentions and Capabilities," November 2007, pp. 5-6, www.dni.gov/press_releases/20071203_release.pdf.

7. Mark Mazzetti, "Intelligence Chief Cites Qaeda Threat to U.S.," The New York Times, February 6, 2008, www.nytimes.com/2008/02/06/washington/06intel.html?_r=2&n=Top/Reference/Times%20Topics/People/Z/Zawahri,%20Ayman%20Al-.

8. "U.S. Draft Text for Fissile Material Cutoff Treaty," May 18, 2006, http://geneva.usmission.gov/Press2006/0518DraftFMCT.html.

9. Christopher A. Ford, "NPT Article IV: Peaceful Uses of Nuclear Energy," remarks to the 2005 NPT Review Conference, May 18, 2005, www.state.gov/t/vci/rls/rm/46604.htm.

10. See Suzanna van Moyland, "The IAEA's Safeguards Program '93+2': Progress and Challenges," Disarmament Diplomacy, No.11 (December 1996).

11. In 2005, after two years of work to detail Iran's covert nuclear program, for instance, the director-general called on Iran to provide cooperation and transparency above and beyond that required by the Model Additional Protocol. IAEA, "Implementation of the NPT Safeguards Agreement in the Islamic Republic of Iran," GOV/2005/67, September 2, 2005, para. 50 ("such transparency measures should extend beyond the formal requirements of the Safeguards Agreement and Additional Protocol"). See IAEA Board of Governors, "Implementation of the NPT Safeguards Agreement in the Islamic Republic of Iran," GOV/2006/14, February 4, 2006, op. para. 1 (deeming it necessary for Iran to "implement transparency measures...which extend beyond the formal requirements of the Safeguards Agreement and Additional Protocol"). Another aspect of the agency's authorities is that they require some nexus to nuclear material, meaning that most aspects of nuclear weaponization remain outside the agency's mandate to investigate, absent the cooperation of the suspected violator, even if the agency is in receipt of evidence that such activities are ongoing. See IAEA Board of Governors, "Implementation of the NPT Safeguards Agreement in the Islamic Republic of Iran," GOV/2005/67, September 2, 2005, para. 49 (noting that the IAEA's "legal authority to pursue the verification of possible nuclear weapons related activity is limited").

12. Indeed, the Model Additional Protocol was explicitly expected not to be required for the nuclear-weapon states. Had the situation been otherwise, its provisions would of necessity been far less intrusive. Although the United States has adopted the Model Additional Protocol in its entirety, on the important condition that it be able to apply a national security exclusion where needed, the protocols adopted by the other nuclear-weapon states have varying degrees of applicability. In no case do they match the Model Additional Protocol. In effect, this is a concrete example of the argument I suggest about an FMCT. Making the Model Additional Protocol universal and nondiscriminatory would have kept it from being particularly useful in helping prevent nuclear weapons proliferation: it could either include authorities that would help with the detection of undeclared activities in non-nuclear-weapon states or it could apply equally to all. It could not do both.

13. For a discussion of recent NSG discussions, see Miles A. Pomper, "Nuclear Suppliers Make Progress on New Rules," Arms Control Today, December 2008, p. 52.

14. The term "nondiscrimination" is unfortunate. Powerfully evocative of generations of legally sanctioned racial and religious prejudice in countries around the globe, the phrase is usually taken in its diplomatic context to mean that all countries should be treated identically. The NPT, for example, is said to be discriminatory because it has one set of rules for recognized possessors of nuclear weapons and one set for non-nuclear-weapon states. It is a category error, however, to assume that treating different countries differently should always be offensive in the way that is racial discrimination. After all, racial prejudice offends our morals because it treats people as being fundamentally different who are in fact fundamentally alike. By contrast, treating a state that possesses nuclear weapons differently than a nonpossessor for purposes of regulating nuclear technology within a nonproliferation regime is not to commit discrimination but instead to recognize and act on a contextually meaningful difference. To refuse to treat unlike things differently is not being nondiscriminatory, it is being indiscriminate, which in other contexts, at least, is often reckoned crude and foolish. In some circumstances, one-size-fits-all rules are not very wise.

15. IAEA, "The Structure and Content of Agreements Between the Agency and States Required in Connection With the Treaty on the Non-Proliferation of Nuclear Weapons," INFCIRC/153 (Corrected), June 1972, para. 14.

16. Brazil, a country that entertained nuclear weapons ambitions in the past, today claims to be pursuing a nuclear submarine program and already conducts uranium enrichment at its Resende facility under conditions partially concealed from IAEA inspectors. One might wonder whether this is something that supporters of the safeguards system really wish to encourage or to afford a patina of legitimacy under paragraph 14.

17. Treaty on the Non-Proliferation of Nuclear Weapons, art. II (hereinafter NPT). The treaty opened for signature on July 1, 1968, and entered into force on March 5, 1970.

18. See U.S. Department of State, "Adherence to and Compliance With Arms Control, Nonproliferation, and Disarmament Agreements and Commitments," August 2005, pp. 64-65, www.state.gov/documents/organization/52113.pdf.

19. NPT, art. IX(3) ("For the purposes of this Treaty, a nuclear-weapon State is one which has manufactured and exploded a nuclear weapon or other nuclear explosive device prior to 1 January 1967.")

20. This discussion assumes that the right number, in this regard, is indeed eight and not nine. North Korea has promised to abandon its nuclear programs and return to the NPT as a non-nuclear-weapon state and should be held to this commitment.

21. As articulated by U.S. negotiator Paul Nitze during the negotiations with the Soviet Union on the Intermediate Nuclear Forces (INF) Treaty during the 1980s, "effective verification" meant in part that "if the other side moves beyond the limits of the treaty in any militarily significant way, we would be able to detect such violation in time to respond effectively and thereby deny the other side the benefit of the violation." During the negotiations with Russia that led to START, Secretary of State James Baker added the qualification that effective verification must also be able to detect "patterns of marginal violations that do not present immediate risk to U.S. security." David Hafemeister, "The Comprehensive Test Ban Treaty: Effectively Verifiable," Arms Control Today, October 2008, p. 6-7.

22. A contemporary example of the importance of this last factor can be found in the Strategic Offensive Reductions Treaty (SORT). Despite their general inability to confirm specific Russian warhead levels, U.S. officials found additional verification mechanisms to be unnecessary in part because they assessed that Moscow could not maintain Cold War warhead stocks anyway and wished for its own reasons to reduce Russia's holdings more or less to the very levels contemplated by the treaty. There was felt to be no particular need to worry about significant levels of cheating under the SORT.

23. Ironically, even this qualification would not apply at very low nuclear-weapon-state warhead levels, when it would presumably become impossible to live with the unavoidably significant FMCT verification uncertainties because any production of fissile material for weapons purposes anywhere could represent a grave challenge to the global balance of power. At the point of a disarmament "zero," of course, this problem is the most acute. The more progress is made toward nuclear disarmament, therefore, the more untenable an FMCT verification regime would become. This is an ironic point apparently not recognized by many disarmament advocates.

The context of a U.S.-Soviet (or U.S.-Russian) nuclear balance with large numbers of warheads and delivery systems may have allowed effective verifiability to be more easily achieved in past superpower arms control negotiations than it is becoming in the international community's post-Cold War struggles with issues of nonproliferation and disarmament. Applying the Nitze-Baker "military significance" test in the nuclear weapons proliferation context or, rather more hypothetically, with regard to issues of nuclear disarmament at very small arsenal levels and especially at zero would surely be particularly demanding. As noted above, uncertainty about the possible existence of an extra handful of weapons here or there might perhaps have been acceptable in the context of a Cold War nuclear standoff between parties already possessing many thousands of such devices. The threshold of military significance arrives much more quickly where at issue is the potential arrival of a completely new player in the nuclear weapons business or one country's achievement of breakout from a nuclear weapons abolition regime. There can be very little margin for error in these latter cases, precisely because preventing acquisition of the marginal weapon is the whole point of the exercise. For today's nuclear nonproliferation and disarmament challenges, therefore, one should not be surprised to find decision-makers frequently needing more in order to achieve effective verification than was necessary for Cold War arms control.