- Jay Davis
- David Kay
- John Lauder
- William Tobey
Moderated by the Wisconsin Project on Nuclear Arms Control
As nuclear negotiations with Iran near their final stage, the question of inspections has come to the fore. If a final agreement is reached, inspections will be a principal means of assuring that Iran does not develop nuclear weapons, either by “breakout” at declared facilities, or by “sneakout” using secret sites.
Given the importance being placed on inspections, what type will be necessary? What inspection and verification regime will be needed to facilitate compliance, detect violations, and ensure effective enforcement? Experience with past inspections can inform this debate. For example, in both Iraq and North Korea, inspections achieved some success but were never able to provide a full understanding of the programs being inspected. That was true in Iraq despite years of effort by the most intrusive, well-equipped, and fully empowered inspection teams ever fielded. Can we do better in Iran? If so, what would be needed by way of authority, enforcement, and operational capability?
These questions were addressed at a private roundtable discussion moderated by Valerie Lincy, Executive Director of the Wisconsin Project on Nuclear Arms Control, in Washington, D.C. on April 23, 2015. The participants were Jay Davis, Senior Fellow and President Emeritus of the Hertz Foundation; David Kay, Senior Fellow at the Potomac Institute for Policy Studies; John Lauder, an independent consultant on nonproliferation and arms control; and William Tobey, Senior Fellow at Harvard University’s Belfer Center for Science and International Affairs.
Highlights of the Findings
There was broad consensus that a final deal should require Iran to provide a declaration of its past and present nuclear work. To verify the declaration, inspectors must be allowed to mount “challenge inspections” of suspicious sites – including military ones, a principle Iran has accepted in other arms control agreements. Inspections may be led by the International Atomic Energy Agency, but the P5+1 countries should remain responsible for their success. Inspection teams should be augmented by expert personnel from the P5+1, be allowed to use the latest inspection equipment, and to cooperate with national intelligence services. Should cheating occur, inspectors are most likely to encounter an accumulation of ambiguities or small violations, rather than dramatic evidence of illicit activity. The usually suggested response – sanctions – will be difficult to re-impose, especially toward the latter part of the agreement, when most sanctions will have been lifted.
Following is a summary of the discussion, expressed as findings. The findings are a composite of the panelists’ individual views. No finding should be attributed to any single panelist or be seen as a statement of the policy of any government.
A final deal should require Iran to provide a declaration of its past and present nuclear work, which is essential for carrying out effective monitoring and verification.
Declaration: A full declaration of Iran's past and present nuclear work is an essential baseline for inspections and for determining Iran’s commitment to compliance. There are four reasons why such a declaration is needed:
- The declaration’s content – and any omissions – would tell inspectors whether Iran is accepting monitoring and verification in an honest way. It could assist inspectors in determining whether Iran is withholding information and would increase confidence in a final agreement.
- A full declaration provides the essential starting point for both monitoring and verification. A declaration shows inspectors where to begin looking and alerts them to sites of possible concern. Without a declaration that inspectors can rely upon as complete and accurate, inspectors would be placed in the untenable “cat and mouse” scenario they faced in Iraq, where they had to look everywhere to uncover undeclared work. Such a task strains the resources of inspection teams and creates a climate of confrontation. Without trust in the completeness and accuracy of the declaration, the inspectors would face “an almost impossible inspection task,” in the words of one panelist.
- A full declaration provides the basis for an ongoing dialogue with Iran. It is an important departure point in enhancing communication with Iran at the working level and a method for facilitating more open and transparent Iranian behavior. The declaration can be reviewed by national intelligence agencies and other organizations, and any ambiguities can be raised and discussed in a “Joint Commission,” such as the one described by the November 2013 Joint Plan of Action (JPOA).
- A full declaration provides assurance to third-party countries and the general public. Unlike national intelligence reports or classified annexes, a declaration would be shareable with key allies like Saudi Arabia and Israel – concerned countries not a part of the P5+1 talks and outside a Joint Commission. The declaration could also be shared with and scrutinized by private analysts and the public at large.
The panel concluded that the declaration need not be a “confession” of illegal activity, but would need to include all prior relevant nuclear work. The scope would be more limited than what was required from Iraq in U.N. Security Council resolution 687. But it would need to provide enough data to construct a baseline of Iranian activity and to conduct effective inspections.
The panel expressed concern that no firm requirement for Iran to make a declaration has been agreed in the negotiations, judging from the information about the framework agreement released so far. Such a requirement must be established. In an effort to avoid controversy, the declaration could take the form of a consolidation and updating of the various declarations that Iran has made since 2003. Iran provided a written declaration of its past nuclear activities to the International Atomic Energy Agency (IAEA) in 2003 and has submitted some updated information since then. It also made its initial declaration under the Agency’s Additional Protocol in 2004 and has provided additional declarations as required by the JPOA. And under a final agreement, Iran would implement the Additional Protocol, which requires an “expanded declaration” covering additional nuclear material and facilities, as well as nuclear-related imports.
“Possible military dimensions” (PMD): The U.S. government, in its April 2 fact sheet describing the framework of a final agreement, states that “Iran will implement an agreed set of measures to address the IAEA’s concerns regarding the Possible Military Dimensions (PMD) of its program.” This suggests that the P5+1 intends to leave the PMD issue to be settled by Iran and the IAEA. The panelists agreed that Iran must fully address and resolve the IAEA's questions about its past military work on its nuclear program. Several panelists noted that a declaration from Iran about its prior military work would provide the basis for inspectors to seek access to military sites where nuclear work occurred. Without such a declaration, the IAEA's authority to access military sites – and other sites not explicitly part of Iran's nuclear program – would be diminished. Panelists also agreed that a final deal should not undercut the work of the IAEA, which has raised the PMD issue more insistently than has the United States government.
A final deal should authorize a challenge inspection of any facility or location in Iran.
To allow the IAEA to visit suspicious, undeclared facilities – including military sites – a final deal should establish an inspection regime based on a principle Iran has already accepted in another arms control agreement: “challenge inspections.”
Under the terms of the Chemical Weapons Convention (CWC), to which Iran is a party, any member-state can call for an on-site challenge inspection of any facility or location in the territory of another member-state. The challenged state cannot refuse access or delay the inspection, but under the principle of “managed access,” it enters into negotiations with inspectors over the extent of access to be granted. Managed access procedures allow the challenged state to protect sensitive information not related to chemical weapons yet are designed to provide inspectors access to all the information that is relevant to their inquiry. Iran has also agreed, in the JPOA, to provide IAEA inspectors with managed access to its uranium mines and mills, centrifuge assembly workshops, centrifuge rotor production workshops, and centrifuge storage facilities. In addition, the IAEA’s Additional Protocol allows for “complementary access” to undeclared sites designated by the Agency, in order to resolve inconsistencies or questions.
A final agreement should enhance the IAEA’s authority. It should, at a minimum, enable the IAEA to investigate undeclared sites in Iran, on a challenge basis, under the principle of managed access. The inspectors should also be given ready access to related documents, facility plans, and key personnel, which the panel deemed essential. The modalities for a challenge inspection regime should be established immediately, perhaps through something resembling an inspection exercise. Both Iran and the inspectors could run through the requirements related to speed and breadth of access, provision of supporting documents, and interviews with relevant personnel.
The concept of “anytime, anywhere” inspections, although used in connection with challenge inspections, is not quite the same. The former was granted to inspectors in Iraq who then claimed the right “to carry out inspections, at any time and without hindrance, of any site, facility, area, location, activity, material or other item,” with the exception of designated presidential sites. Regardless of whether such authority is granted to inspectors in Iran, inspectors must always depend on the host-country for security and transportation and will never be free simply to roam about and show up unannounced at a suspect site. Their access to sites will always be managed to some degree.
In addition, “anytime, anywhere” inspections should not be considered a panacea, or a guarantee of a robust monitoring and verification regime. Inspections are the last step in the monitoring process, not the first. Far more important than unrestricted access to sites will be the accuracy of Iran’s declaration, full access to plans and records describing key sites, and conversations with key personnel – all of the data that would show inspectors whether and where they should launch a challenge inspection. The case of Kumchang-ni in North Korea offers a cautionary tale. Based on intelligence reports of a secret, underground nuclear facility at Kumchang-ni, the United States launched protracted, months-long negotiations with the North Koreans in 1998 to gain access. Yet after finally getting into the site in 1999, inspectors found no evidence of nuclear activity.
National intelligence services will be a vital asset for effective monitoring and verification; however, supplying intelligence information to inspectors could prove difficult.
Intelligence information will play a vital role in inspections in Iran. Information gathered by national intelligence services will help inspectors decide where to look, what to look for, and how to evaluate what they find. In Iraq in particular, intelligence support was crucial to inspections. An information loop was established in which intelligence agencies provided information to inspectors, and inspectors provided the results of inspections back to intelligence agencies for analysis, which led to further inspections. For that reason, a final agreement should authorize cooperation between the inspectors and national intelligence agencies.
Yet this cooperation may be difficult. Sensitive information that the United States gathers, for example, will be difficult to share with the other P5+1 countries. The United States has a general intelligence sharing agreement with only one P5+1 country – the United Kingdom – and has an adversarial relationship with two others. The IAEA has mechanisms that allow member states to pass information to the Agency, which have been used by the United States and other countries. There is precedent for this in Iran. An IAEA team devoted to Iran has cooperated with national intelligence services. The final agreement should establish a commitment on the part of the P5+1 and the IAEA for intelligence information sharing to facilitate implementation of a final agreement. Still, a feedback loop like the one that existed in Iraq may be difficult to achieve, particularly if the inspection teams do not include U.S. citizens. Sharing certain nuclear information with a non-U.S. citizen also could run afoul of the U.S. Atomic Energy Act.
A final deal must allow IAEA inspection teams to be augmented by expert personnel, and to use the latest inspection equipment.
A final agreement should establish the authority of the P5+1 to augment IAEA inspections teams with expertise the IAEA may not possess internally. Inspection teams need expertise, for example, in analyzing facilities to determine whether they are being used for their ostensible purpose. Nuclear weapon experts and experts in centrifuge manufacturing and production could also enhance IAEA teams. Augmenting inspection teams with officials from P5+1 countries could also mitigate difficulties related to intelligence sharing described above.
There is also the fact that the IAEA does not have a counterintelligence capability. During the course of a ten or fifteen year agreement, Iran will attempt to gain access to inspection planning at numerous levels: at IAEA headquarters, on inspection teams in-country, and at the United Nations. Iraqi agents accomplished this in the 1990s. They operated in New York, Vienna, and in Iraq itself. Robust counterintelligence will be indispensable to the integrity of inspections in Iran. The IAEA must be equipped with experts that will give it and its inspection teams counterintelligence capability.
There is precedent for such augmentation. The IAEA has fielded teams with augmented personnel in Iraq and South Africa. In order to ensure that inspection teams have the requisite expertise, Iran should not be allowed to refuse entry or a visa to any inspector.
In addition to enhanced expertise, the inspection teams should benefit from the latest inspection equipment. And they should be allowed to upgrade their equipment throughout the duration of a final agreement. This authority is mentioned in the April 2 joint statement released by the European Union and Iran, which stated tersely that the IAEA “will be permitted the use of modern technologies.” Inspectors should be free to choose, in particular, off-the-shelf, open-source equipment instead of being limited to equipment that is “black boxed.” The latter could be seen as a means of spying and may produce information difficult to share with other countries. It would be good to have “device-level” telemetry. Iran, in fact, may be amenable to the use of more efficient equipment because it could reduce the need for on-site inspectors, who are perceived as more intrusive.
The panel agreed that decisions about specific monitoring technologies could be the task of a Joint Commission, rather than a bilateral arrangement between the IAEA and Iran. This has the benefit of including the P5+1 without excluding Iran.
The panel also agreed that the IAEA should be responsible for all inspections in Iran directly on behalf of the P5+1, and that this authority be set forth in a U.N. Security Council resolution. Such a resolution would provide special status and access to the P5+1 for IAEA inspectors.
A final deal must enable speedy reporting of violations, and a means for resolving disputes.
The P5+1 countries will remain responsible for seeing that any agreement with Iran is carried out. The P5+1, together with the IAEA, must decide how to structure, organize, train, and equip inspection teams, and how to handle their reports. A final deal should establish a reporting and governance structure, centered on a Joint Commission, with the IAEA reporting directly to this consultative body, and to the U. N. Security Council. During the inspections in Iraq after the first Gulf War, the IAEA was able to conduct its work expeditiously because it reported directly to the U.N. Security Council and the U.N. Special Commission on Iraq (UNSCOM) rather than to its own Board of Governors. A similarly expeditious reporting scheme should be established for the inspections in Iran.
The P5+1 should also empower a Joint Commission to mediate disputes and resolve ambiguities that could undermine an agreement over the long-term. The panel emphasized that a consultative commission of some kind has been a feature of every successful arms control agreement.
Judging from its pattern of behavior over the past decade, Iran probably will not commit a glaring violation of its commitments. Rather, the P5+1 countries may encounter inconsistencies, ambiguities, or selective opacity from the Iranians. Resolving these ambiguities expeditiously, through an established dispute resolution process, is both necessary for the inspection process, and to show inspectors that their work is not being ignored. It would also reassure countries in the region that the agreement is being carried out.
A Joint Commission should operate on two levels. A technical level should address the day-to-day issues of compliance at the working level – independent of political influence – while a political level would take up any questions that cannot be resolved at the technical level. Also, no single country should have a veto in a Joint Commission.
If export controls are relaxed on nuclear and missile-related items, Iran must be required to make all such imports through an established procurement channel. Inspectors should be authorized to monitor and verify the end-use of any such item.
The panel agreed that all items “dual-use and above” should be imported only through an official procurement channel, and that the end-use of all procured items, including use in sectors outside Iran's declared nuclear program, should be subject to verification. This should be a relatively straightforward task if inspectors are given sufficient authority through a U.N. Security Council resolution. It would also provide inspectors with an additional basis for accessing undeclared facilities. Verifying the procurement channel will be especially important for Iran's centrifuge program, which will need imports for research and development. If Iran's centrifuge effort increases in the later years of an agreement, it would become more important for inspectors to see the kinds and volumes of items procured. This knowledge would allow them to assess Iranian capabilities and intentions.
In addition, the channel would create a bright line between legitimate and illicit trade. Any procurement of controlled items outside the channel would be considered a violation and would be investigated by inspectors.
Despite its importance, this procurement channel may be difficult to oversee. The precedent is the “oil-for-food” program in Iraq. Although the program prevented Saddam Hussein from rebuilding his army, it proved to be a vehicle for massive international corruption. The P5+1 should attempt to avoid such an outcome for Iran.
Over the term of a ten or fifteen year agreement, the ability to enforce inspection obligations will inevitably decrease.
In the implementation of an inspection agreement, sanctions should be considered a risk-reward mechanism. The gradual lifting of sanctions on Iran should be a confidence-building measure to reward compliance with the agreement, while the re-imposition of sanctions should be a punitive measure to deter or punish non-compliance.
As stated above, the P5+1 is likely to encounter numerous small violations rather than a single egregious one. Yet, unlike reciprocal arms control agreements (such as the Cold War accords between the United States and the Soviet Union), there is no symmetry between these small violations and the probable options for response. The Iran deal is more like a “cash for access” arrangement. If the bulk of sanctions are lifted early, then most of the “cash” has already been given. This reduces Iran’s incentive to grant access to inspectors and leaves only a frail instrument – the re-imposition of sanctions when most have already been lifted – to punish violations. In Iraq, by comparison, sanctions were meant to stay in place until inspectors verified disarmament. Although military action is another option, it too would be seen as disproportionate to small violations.
Multilateral sanctions – those that have had the greatest impact – will not be easy to re-impose, even with general agreement that such action is necessary in response to violations. The characterization of the re-imposition of sanctions as “snapback” is a misnomer, the panel concluded. It took about seven years of diplomacy to achieve the present level of sanctions; reinstating them, especially at the United Nations, will not be expeditious. Russia and China are unlikely to give up their veto authority and accept automatic sanctions triggers. European companies may have renewed commercial ties and may not want to give them up. They will lobby their governments strenuously to oppose losing either markets or investments. European Union sanctions, in particular, will require support from 27 countries to re-instate. And the United States is unlikely to use secondary sanctions to punish European firms that do business in Iran.
In addition, the duration of a long-term deal – ten to fifteen years – will pose a monitoring and verification challenge. Over time, countries will become reluctant to recognize violations, or investigate controversial inconsistencies, because they will not want the deal to fail. Governments will have a vested interest in success and will resist bad news. This is a traditional problem in long-running arms control agreements. The effect is to discourage inspectors, who will lose their incentive to report violations. “Inspection fatigue” will accumulate and the desire to investigate small violations or inconsistencies will diminish, especially if there is no follow-through at higher levels of the IAEA or P5+1 countries. Inspectors, who are often tasked with mundane, repetitive, and indeed boring monitoring tasks, may over time become accustomed to seeing inconsistencies as “normal” and not flag them for further investigation.