The Irreversibility Paradox: What Makes for Enduring Arms Control and Disarmament

ABSTRACT

The principle of irreversibility poses a paradox for arms control. On the one hand, negotiators often seek to ensure that progress made in nuclear arms control is enduring and that agreements are resilient over time. This may include the dismantlement of certain types of weapons or the elimination of weapons-usable fissile materials. Simultaneously, policymakers must seek to ensure that arms control agreements are palpable to domestic stakeholders, especially if the treaties require legislative consent, and have the flexibility to respond to changes in the security environment. This often takes the form of withdrawal clauses, giving parties the option to leave an agreement and reverse any potential gains or the underlying intent of an agreement. This paradox bears many similarities to the transparency-security tradeoff identified by Coe and Vaynman whereby, “Any deal that is transparent enough to assure that one side complies with the deal may also shift the balance of power so much that the other side reneges in order to exploit the shift”. We argue that a parallel tradeoff is between political, legal, and technical measures, including verification and transparency, which can confirm irreversibility, and security concerns that will motivate states to build flexibility into agreements, making them less irreversible, i.e. more reversible.

KEYWORDS:

• Irreversibility
• nuclear disarmament
• negotiations
• nuclear deterrence
• arms control

Introduction

The principle of irreversibility poses a paradox for arms control. On the one hand, negotiators often seek to ensure that progress made in nuclear arms control is enduring and that agreements are resilient over time. This may include the dismantlement of certain types of weapons or the elimination of weapons-usable fissile materials. Simultaneously, policymakers must seek to ensure that arms control agreements are palpable to domestic stakeholders, especially if the treaties require legislative consent, and have the flexibility to respond to changes in the security environment. This often takes the form of withdrawal clauses, giving parties the option to leave an agreement and potentially reverse any gains or the underlying intent of an agreement. This paradox bears many similarities to the transparency-security tradeoff identified by Andrew Coe and Jane Vaynman whereby, “Any deal that is transparent enough to assure that one side complies with the deal may also shift the balance of power so much that the other side reneges in order to exploit the shift” (Coe and Vaynman 2019). We argue that a parallel tradeoff is between political, legal, and technical measures, including verification and transparency, which can strengthen and confirm irreversibility, and security concerns that will motivate states to build flexibility into agreements, making them less irreversible.

What makes some arms control agreements more irreversible than others? How do states reconcile the irreversibility paradox in negotiating arms control agreements? The concept of irreversibility is not new. In the context of the Treaty on the Non-Proliferation of Nuclear Weapons (NPT), irreversibility has been referenced in many international forums in relation to nuclear disarmament, often in tandem with the concepts of verification and transparency. The first reference in the context of the NPT came from the 2000 NPT Review Conference Final Document, which outlined irreversibility as a principle both in eliminating weapons-grade fissile material stocks and as a general principle in nuclear disarmament (United Nations 2000). Despite now being used as common language in diplomatic negotiations and discussions of nuclear disarmament, little work has been done to understand the concept of irreversibility, particularly its successes and challenges in the context of arms control and what this might suggest about irreversibility for nuclear disarmament (IND).

There is a growing conventional wisdom that irreversibility should be thought of as a spectrum, rather than as a binary (Cliff 2011). For the purposes of this paper, irreversibility is defined as a quality of a disarmament or arms control process that involves limiting the capacity for re-armament, including the possible re-constitution of aspects of weapons programs. The International Partnership for Nuclear Disarmament Verification (IPNDV) evaluates irreversibility in terms of effectiveness – referring to the level of confidence in compliance – and verification across political, legal, and physical measures (International Partnership on Nuclear Disarmament Verification 2018, 8–9). Several authors identify transparency and verification measures as critical to building observable irreversibility. Bunn and the IPNDV present observability as inherent to the definition of irreversibility (International Partnership on Nuclear Disarmament Verification 2018, 1–2, 8). The IPNDV argues that “adequate irreversibility” is met “when the country cannot reverse its commitment without the observing country being made aware” (Cliff 2011). In this sense, irreversibility “[relies] on very substantial and robust verification measures” (Cliff 2011, 8). Yet this definition, and others, is at odds with the emerging conventional wisdom in policy and academic literature that envisions irreversibility as a spectrum. If irreversibility is more than just verification and technical factors, as captured in the irreversibility paradox, states will have to make tradeoffs between a variety of other factors. Filling this gap in scholarship on irreversibility will not only contribute to the growing body of literature on disarmament drivers and dynamics (Pelopidas 2015; Ritchie 2022), but also contribute to nuclear diplomacy efforts to bring together nuclear possessors and non-possessors to strengthen the NPT and wider nuclear order (Gwadera 2022).

This article has three main objectives. First, we aim to expand scholarship on irreversibility by analyzing historical cases of arms control to identify how states navigated the irreversibility paradox. Second, we strive to make a conceptual contribution to academic research on irreversibility by identifying factors beyond verification that inform states’ decisions about irreversibility, including where to fall on the irreversibility spectrum. And finally, we aim to offer a future research agenda to inform bridge-building efforts around IND. A diverse group of states have shown interest in the IND project and engaged with it as a meaningful step towards fulfilling NPT Article VI commitments. A 2023 Wilton Park workshop concluded that IND “could be seen as a gradual procedure to build trust and consequences that belong to a broader set of nuclear disarmament instruments and commitments” (Rodriguez 2023).

This article proceeds in three sections. First, we outline a conceptual framework for thinking about the irreversibility paradox, drawing on Coe and Vaynman’s transparency-security tradeoff. Next, we explore six historical cases: the Plutonium Management and Disposition Agreement (PMDA), the Intermediate Range Nuclear Forces Treaty (INF), the Joint Comprehensive Plan of Action (JCPOA), the New START treaty, the Presidential Nuclear Initiatives (PNIs), and the Biological and Toxin Weapons Convention (BWC). And finally, we outline a future research agenda to inform nuclear diplomacy and policy, particularly bridge-building efforts between nuclear possessors and non-possessors. Ultimately, we argue that the irreversibility paradox will present numerous challenges for future efforts at IND but may also offer solutions by finding the right balance of resilience and flexibility. We demonstrate that, somewhat unexpectedly, agreements with more flexibility are the more resilient over time. Our overarching finding is that many of the non-technical factors in irreversibility have been overlooked and should be incorporated into a future research agenda. Perhaps the most important finding of this article is that irreversibility is, above all, a political exercise and the article points to a variety of ways to improve upon the concept of irreversibility as a spectrum by considering temporal and domestic factors.

The Irreversibility Paradox

There is tension in the concept of irreversibility in arms control agreements. On the one hand, states want agreements that stand the test of time. This is particularly true for arms control between adversaries or competitors, such as those during the Cold War between the United States and the Soviet Union. The United States was only willing to accept limits on its strategic capabilities in exchange for assurances that the Soviets would do the same over time. In competitive relationships, states may accept high degrees of irreversibility as long as it is reciprocal.

On the other hand, even in reciprocal agreements, states may be reluctant to constrain themselves and limit their strategic options. This is particularly true in times of high uncertainty and geopolitical instability, such as the current landscape following Russia’s invasion of Ukraine, expansion of China’s strategic capabilities, and rise of authoritarian leaders. States often attempt to manage this uncertainty and give themselves flexibility by setting time limits on arms control agreements, such as the Strategic Arms Limitation Talks Interim Agreement (SALT I), which lasted for five years, or the New Strategic Arms Limitation Treaty (New START), which is set to expire in 2026. The result of these two competing priorities is that states must negotiate some degree of flexibility and reversibility into arms control.

Coe and Vaynman present a similar dilemma in their security and transparency tradeoff (Coe and Vaynman 2019, 342–345). Arms control agreements with too much transparency may reveal unintended information about a state’s military capabilities. However, states need some degree of transparency to ensure that the other side is cooperating and not undermining the security agreement. Coe and Vaynman demonstrate how this tradeoff undermined several attempts at arms control agreements. They describe the tradeoff as follows:

To be viable, any deal must satisfy a transparency requirement and a security requirement. To ensure the compliance of the side that could arm, the probability that its cheating will be detected must be high enough: monitoring must render this side’s arming sufficiently transparent. However, the information revealed by monitoring also cannot give the monitoring side too large a military advantage: that is, the deal must be secure for the arming side … . The problem is that transparency may reveal not only a state’s arming decision but also other information relevant to the balance of power. For example, inspections intended to confirm that a military base does not contain prohibited weapons may also yield information that makes it easier to effectively attack the base or enables the inspecting side to assess other military capabilities. Thus, there may be a trade-off between transparency and security.

(Coe and Vaynman 2019, 342–343)

The irreversibility paradox similarly resembles this tradeoff. To ensure compliance with a disarmament or arms control agreement, states will want a high degree of verification and monitoring. However, complying with these requirements may impact states’ ability to respond to changes in the security environment thereby jeopardizing their security over time. In short, irreversibility undermines a state’s ability to hedge. And whether or not states enter into highly irreversible agreements, i.e. how they navigate the irreversibility paradox, will be largely determined by their expectations about the future geopolitical landscape and future security requirements.

There is a growing body of scholarship on how states navigate paradoxes of transparency (Coe and Vaynman 2019; Vaynman 2021), trust (Wheeler 2018), and strategic competition (Williams 2019) in arms control. The paradox of irreversibility has received less attention. The majority of scholarship on IND categorizes irreversibility through three lenses: technical, legal, and political challenges. Many scholars focus on the importance of verification across all three lenses. Persbo argues that verification measures are needed to make rearmament “both time-consuming and costly”, or, difficult to reverse.¹⁰ Others note that transparency and verification are complementary to irreversibility but not inherent to it. Cliff et al. emphasize that “verification does not by any means ensure irreversibility. Nor is it designed to” (Cliff 2011). Instead, transparency and verification can foster credible and timely assurance of compliance with the principle of irreversibility, both during and following the process of disarmament (Anthony 2011, 9; Cliff 2011, 14; Sagan 2010, 166). By fostering this sense of reassurance, transparency and verification could facilitate “oversight of the irreversibility of reductions” and “help to pave the way for nuclear disarmament (Zarimpas 2003, 253)”.

Another prominent approach to irreversibility is to view it as a spectrum (Cliff 2011). At one end of the spectrum, states may be considered disarmed if they maintained production capabilities but eliminated all weapons. If all production capabilities and weapons have been eliminated, then the state occupies the opposite end of the continuum. One end of the spectrum, for example, might be US policy for nonproliferation agreements with North Korea. The Biden Administration continues to insist on verifiable denuclearization. At the opposite end of the spectrum, Sagan, Persbo, and Cliff et al. acknowledge that a low degree of irreversibility implies that former nuclear-weapon states (NWS) could retain some latency and, if needed, use it to hedge or rearm (Cliff 2011, 12–13, 64–68; Sagan 2010, 166) but this latency could introduce a form of stability between states, yet may simultaneously have a destabilizing effect over time. States may not pursue rearmament because if one state shows indicators of reversal then others would also rush to rearm (Cliff 2011, 12–13; Sagan 2010, 166). Alternatively, maintaining latency could provide a sense of reassurance that enables NWS states to pursue disarmament, knowing that rearmament is possible if the security environment mandates it (Anthony 2011, 1–2; Cliff 2011, 12). Observability is a key to irreversibility in the context of both nonproliferation and strategic stability.

Other recent literature has sought to undermine the role of arms control in achieving nuclear disarmament and would likely challenge the premise of this article. David Mutimer has argued that practitioners of arms control “are not only content to see nuclear weapons remain, but … actively require them (Mutimer 2011, 68)”. Similarly, Kjølv Egeland has noted that nuclear arms control efforts “extrapolating from limited nuclear arms control efforts to a general theory of disarmament should be done with extreme caution” (Egeland 2022). Instead, Egeland suggests looking towards the abolition of morally reprehensible practices such as slavery. Arms control and disarmament are two distinct practices in international affairs; however, they can also be mutually reinforcing. For example, arms control during the Cold War led to an 88% reduction in the American nuclear stockpile (US Department of State 2022). And the majority of nuclear possessor states continue to advocate for a step-by-step approach to disarmament, rooted in verifiable and reciprocal arms control agreements. While many scholars may advocate an alternative view of arms control and disarmament, the step-by-step approach remains the one most prominent among the possessor states and therefore is a worthwhile set of cases for understanding when, why, and how states would engage in the practice of irreversibility for disarmament.

In the face of this paradox and competing frameworks of irreversibility, how do states reconcile the paradox in the context of arms control negotiations when they face multiple competing pressures to achieve agreement but without tying their hands in the face of future uncertainty? One of the most obvious ways that reversibility is built into almost all arms control agreements is the withdrawal clause. Interestingly, the clause language is consistent across most arms control agreements. The Partial Test Ban Treaty, the NPT, the INF Treaty, New START, the BWC and several other arms control agreements all contain the same withdrawal clause. The clause states that “[E]ach Party shall, in exercising its national sovereignty have the right to withdraw from this Treaty if it decides that extraordinary events, related to the subject matter of this Treaty, have jeopardized the supreme interests of its country”. This language of withdrawal, often referred to as the Extraordinary Events Clause, allows for a degree of reversibility in nearly all arms control agreements.

Despite subsequent use of the Extraordinary Events Clause, there remain serious legal questions regarding what constitutes an extraordinary event and under what conditions the withdrawal clause can be executed. The Extraordinary Events Clause has been claimed by states a handful of times to justify withdrawal from arms control. Perhaps most notably, North Korea claimed the Extraordinary Events Clause to withdraw from the NPT in 2003. The United States first exercised the Extraordinary Events Clause to withdraw from the Anti-Ballistic Missile Treaty on 13 December 200113 December 2001. The United States explained its withdrawal in the following statement:

Since the Treaty entered into force in 1972, a number of state and non-state entities have acquired or are actively seeking to acquire weapons of mass destruction. It is clear, and has recently been demonstrated, that some of these entities are prepared to employ these weapons against the United States. Moreover, a number of states are developing ballistic missiles, including long-range ballistic missiles, as a means of delivering weapons of mass destruction. These events pose a direct threat to the territory and security of the United States and jeopardize its supreme interests. As a result, the United States has concluded that it must develop, test, and deploy anti-ballistic missile systems for the defense of its national territory, of its forces outside the United States, and of its friends and allies.

This interpretation of the Extraordinary Events Clause cited external factors to justify withdrawal. Another justification for withdrawal using the Extraordinary Events Clause is consistent cheating. The US withdrawal from the INF Treaty in 2019 cited Russian violations of the treaty. The statement claimed, “Russia failed to return to full and verified compliance through the destruction of its noncompliant missile system – the SSC-8 or 9M729 ground-launched, intermediate-range cruise missile” (Pompeo 2019). But Extraordinary Events Clauses and withdrawal options demonstrate that no agreement is totally irreversible. Otherwise, for the sake of national interest, most states would not sign up to the agreement.

Balancing Flexibility and Transparency: Historical Efforts to Avoid Irreversibility

To better understand how states navigate the irreversibility paradox in agreements, we explore six short case studies. The case studies were chosen for variation in membership (bilateral and multilateral), limitations (destruction of weapons, bans), and outcomes (still in effect, expired). The six cases are the Biological and Toxin Weapons Convention (BWC), the Intermediate Range Nuclear Forces Treaty (INF), the Presidential Nuclear Initiatives (PNIs), the Plutonium Management and Disposition Agreement (PMDA), the New START treaty, and the Joint Comprehensive Plan of Action (JCPOA). For each short case study, we provide a background of the arms control agreement then explore how states navigated political, legal, and technical factors before and during the agreement. We also examine whether the agreements endured over time. This temporal aspect is often overlooked in IND literature but was identified as an important factor in the transparency-security tradeoff and in our understanding of the irreversibility paradox as states’ ability to hedge, determined by their vision of the future security environment, informs when they participate in arms control.

We acknowledge this is a broad approach to the case studies; however, we opted for a high degree of variation as a “first cut” at applying case study analysis to irreversibility. The analysis not only points to trends in how states reconcile the irreversibility paradox, but also opportunities for future case study analysis, such as holding constant the parties and limitations, but varying the outcome.¹

BWC

The BWC represents an early attempt to negotiate a multilateral treaty banning the misuse of a dual use technology. Opening for signature in 1972 and entering into force in March of 1975, the BWC bans the stockpiling, development, and acquisition of biological weapons. It supplemented the 1925 Geneva Protocol, which banned the use of biological weapons but not their development. The BWC lacks any formal verification mechanism to monitor compliance. Despite this, there are several confidence-building measures and consultative mechanisms built into the treaty. The BWC has set a norm against the use of biological weapons in warfare. However, the BWC has been riddled with compliance issues from the outset, and the lack of verification mechanisms has caused many to doubt the effectiveness of the BWC.

The BWC has several legal mechanisms for review of compliance issues. Under Article V of the BWC, states undertake to consult bilaterally and multilaterally on solving compliance issues. Additionally, Article XII establishes the 5-year BWC Review Cycle. During this review, it is mandated that states “take into account any new scientific and technological developments relevant to the convention” (Nations 1975). Despite the existence of legal review mechanisms, the products of the review conference meetings are often not legally binding. For example, states parties to the BWC are required to submit Confidence Building Measures (CBMs) documents pursuant to the final document of the Second Review Conference in 1986. These CBMs detail information about research centers involved in biological defense programs, information about outbreaks of disease, and legislation that states have adopted to ensure domestic compliance with the BWC’s central provisions. The CBMs provide less effective reporting metrics because few states submit them. 2022 was the highest year of submissions with only 98 of the 185 states parties to the BWC submitted Confidence Building Measures documentation. In 2003, only 35 countries submitted their documentation (United Nations 2023). Overall, there is a lack of legal force behind the few reporting forms that exist in the BWC.

Almost immediately after the BWC went into force, the United States began alleging that the Soviet Union was in non-compliance. For example, in 1981, the United States claimed that the Soviet Union supplied mycotoxins to communist forces in Laos and Vietnam during the “Yellow Rain” incident. While it is unlikely that the Yellow Rain incident was an act of Soviet biological warfare (Meselson, Matthew, and Jullian 2008), the Soviet Premier Boris Yeltsin admitted in 1992 that it had maintained an active offensive biological weapons program in violation of the BWC since signing the treaty.² Despite Soviet noncompliance, the lack of political interest in the United States to develop a large scale offensive biological weapons program manifest little incentive for the United States to reverse its decision to abide by the BWC.

In 1994, after recognizing that the Soviet Union had a large offensive biological weapons program in violation of the treaty, the international community met to consider how to strengthen the BWC. A group of governmental experts produced the VEREX report, which provided 21 recommendations for how to strengthen the BWC. For decades, the BWC review conferences have attempted to incorporate some recommendations from the VEREX report but have failed to garner the needed international support for creating verification mechanisms in the BWC.

INF Treaty

The INF Treaty was a legally-binding agreement with unprecedented intrusive verification mechanisms, and, as was later proven, would be politically costly to reverse. The INF Treaty was a treaty between the United States and Soviet Union that eliminated intermediate-ranged ground-launched missiles and launchers with a range of between 500–5500 km. Notably, the treaty did not apply to sea or air-based intermediate-range or short-range weapons. Additionally, while the treaty included the word “nuclear”, it banned all ground-launched missiles of intermediate range regardless of what type of warhead they carried.

At the time of negotiation, reversing commitments made under INF would have come with high political costs to both Washington and Moscow. One particularly unique factor about INF was the personal involvement and commitment of the political leadership. If INF fell apart, it would have come at a personal cost to Reagan and Gorbachev and undermined both of their political leadership at the time. Additionally, the treaty was universally well-received and touted by the disarmament community, including within the NPT. Gorbachev stated that the INF treaty was “an unprecedented step in the history of the nuclear age” (Oberdorfer and Cannon 1987). If either side went back on its commitments, it would have undermined their reputations as leaders within the NPT and nuclear order more broadly.

As a legally binding agreement, the INF Treaty would have seemed highly irreversible. Article XIII of the INF treaty established a Special Verification Commission (SVC) to discuss compliance challenges. The SVC was designed to resolve questions related to compliance and improve the viability and effectiveness of the treaty. Article XIII also utilized the Nuclear Risk Reduction Centers to facilitate communication for inspection notices and day to day treaty functions.³ The treaty built in legal mechanisms to facilitate communication, thus reducing the possibility of reversal or withdrawal.The SVC was a mechanism for updating the treaty and ensuring continued salience and relevance. Utilizing the SVC, the INF treaty could update methods and technologies used in inspections.⁴

From a technical perspective, the INF Treaty would also seem highly irreversible because it eliminates an entire class of weapons and includes intrusive verification mechanisms. The INF treaty was the first US-Soviet treaty to employ intrusive monitoring, allowing on-site inspections of selected missile facilities. The agreement also included several data exchange systems such as continuous monitoring outside of missile assembly facilities to ensure that intermediate range missiles were not being built. The verification mechanisms were so extensive that Ronald Reagan famously popularized the phrase “trust but verify” during the signing ceremony of the INF treaty in December 1987.⁵ Despite this robust technical verification, many of INF’s monitoring and verification provisions expired 10 years after the elimination of ground launched missiles were completed, although the treaty would remain in force (Woolf and Podvig 2019, 13).

In the mid-2000s, Russia violated the treaty by developing a nuclear-capable intermediate-range ground-launched cruise missiles. As early as 2008, Russia began testing missiles that the United States believed violated the INF treaty, including the SSC-8 cruise missile. In response, the United States withdrew from the treaty in 2019. Russia also believed that the United States was in violation of the treaty, and Russian officials accused the United States of violating the INF because the Aegis Ashore missile defense system includes the same launchers that the United States uses to launch Tomahawk cruise missiles, although the United States retired to nuclear variant of the Tomahawk in the early 2010s.

PNIs

The PNIs were unilateral, non-legally-binding measures taken by the United States and Soviet Union at the end of the Cold War. The initiatives were pursued in the aftermath of the August coup attempt in Moscow, which highlighted not only the potential vulnerability of Soviet nuclear forces, but also the potential for risks of escalation and misunderstanding during crises. The coup attempt prompted both Bush and Gorbachev to make unilateral commitments to reduce non-strategic nuclear forces, such as withdrawal of ground-launched short-range weapons, and end deployment of many naval nuclear capabilities on the U.S. side, and the elimination of nuclear artillery, removal of tactical nuclear weapons from some naval forces, and separating nuclear weapons for missile defense on the Soviet side. One particularly unique trait of the PNIs was the speed with which they were agreed to internally on the U.S. side, driven by Presidential leadership and commitment, with minimal domestic political debate or dissent.⁶

While the agreements were reciprocated, they were not symmetrical as both sides made different types of commitments, and the agreements were not codified in a treaty or other legally-binding agreement. Additionally, they did not include any verification mechanisms. There were some political factors incentivizing the United States and Soviet Union, later Russia, to continue to abide by the agreement, but there were no clear threats of repercussions for any violations of the agreements. The PNIs proved successful in immediately reducing nuclear risks and were observed for decades, but as Russia expanded its arsenal in the 2000s it proved unconstrained by the PNIs and in 2023, and earlier, the State Department evaluated Russia’s nuclear forces as “inconsistent” with its PNI commitments (US Department of State 2023, 5).

PMDA

In the aftermath of the Cold War, several fissile material agreements were negotiated by the Russian Federation and the United States. The PMDA was an agreement to dispose of 34 metric tons of weapons-grade plutonium deemed excess to military needs. If this agreement had been fully completed, it would have eliminated enough weapons-grade plutonium to assemble 17,000 nuclear weapons based on IAEA significant figures (Nikitin 2016). As part of this agreement, both the US and Russia decided to convert excess plutonium into mixed oxide (Mox) fuel to power nuclear reactors. Russia used Mox fuel to power fast-neutron BN reactors. Protocols were put in place to ensure that these reactors did not produce new weapons-grade plutonium (US Department of State 2010b). Russia constructed a Mox fuel fabrication facility that is based in Mayak, which has been online since 2014. The United States also built a Mox fuel facility to dispose of excess weapons-grade plutonium, but the project was never completed. President Barack Obama canceled the construction of the facility in 2016 due to drastic cost overruns and planned to shift to a “dilute and dispose” approach, whereby plutonium would be down-blended with other materials and then buried or crystalized in a process called vitrification.

The PMDA eliminated weapons-grade plutonium in a manner that was comparatively difficult to undo. A factsheet issued by the Department of State in 2010 states that the PMDA “makes arms reductions irreversible by ensuring that the United States and Russia will transparently dispose weapon-grade plutonium from their respective defense programs, thereby preventing the plutonium from ever being reused for weapons or any other military purpose” (US Department of State 2010a).

Ultimately, the Russian Federation unilaterally suspended PMDA. A Russian Foreign Ministry Spokesperson cited the “United States’ obvious inability and unwillingness to honor its obligations … ” as a justification for the withdrawal (American Journal of Law 2017). President Obama declared that the United States was willing to continue to abide by its obligations, and money has been allocated in the U.S. Department of Energy’s budget to pursue and research dilute and dispose methods. However, the revival of the PMDA at a formal level remains uncertain. Russia has repeatedly criticized American plans to dilute and dispose of plutonium, claiming that the plutonium could- in theory- be recovered (Center for Arms Control and Nonproliferation 2021). If the plutonium is buried, it would be possible to dig it up. If the plutonium is vitrified, recovering the plutonium would be costly but still possible. In short, Russian criticisms of dilute and dispose center on the reversibility of the dilute and dispose route.

The PMDA agreement is a demonstration of how attempts at irreversible nuclear disarmament can fail at a technical level even when the scientific knowledge of a process exists. Mox fuel fabrication is technically possible, and several countries have successfully implemented Mox fuel in their fuel cycles. However, the US struggled to implement Mox production due to contractor challenges, cost overrides, and mismanagement. In short, solutions to technical challenges are a necessary but insufficient prerequisite for successful attempts at irreversibility in nuclear disarmament because the process of how technical steps are executed at the project management level is important.

New START

Turning to a case from strategic bilateral arms control, New START limits both the US and Russia to maintain a maximum of 1,550 deployed strategic-range nuclear weapons, and is set to expire in 2026. Despite this, the treaty contains complex counting rules that allow each country to develop its force structure in unique ways. As part of the 2009 reset in US-Russia relations, New START had a high degree of political irreversibility, particularly for the United States. New START is the largest bilateral nuclear arms reduction treaty and has conducted over 300 on-site inspections and over 24,000 notifications. The treaty has eliminated hundreds of nuclear weapons and allows both the United States and Russia to maintain communications through a variety of channels, including the National and Nuclear Risk Reduction Center. Under New START, each nuclear-capable bomber counts as only one warhead despite the number of weapons the airplane is carrying. Because of this, it is technically possible for both the United States and Russia to exceed the 1,550 nuclear weapons by uploading a larger number of nuclear weapons onto the air leg. Each bomber can hold a maximum load of between 16 and 20 nuclear weapons (Arms Control Association 2023).

New START also allows the United States and Russia a higher degree of flexibility than past agreements in how they structure their strategic forces. Additionally, New START, unlike INF, has an expiration date, another mechanism for states to balance and reconcile the irreversibility paradox and have more flexibility in the event of geopolitical shifts. Indeed, flexibility and reversibility within New START proved to be a priority in US domestic debates. The US Senate advised for the ratification of New START on 22 December 201022 December 2010. The vote consisted of 71 in favor, 26 against, and 3 abstentions. All 56 democrats voted supporting the treaty ratification (Senate 2010).

Following the negotiations of New START, Russia made a statement about potential withdrawal because of future US developments in ballistic missile defense. Russia claimed that “the treaty may be effective and viable only in conditions where there is no qualitative or quantitative build-up in the missile defense system capabilities of the United States of America. Consequently, the extraordinary events referred to in Article XIV of the Treaty also include a build-up in the missile defense system capabilities of the United States such that it would give rise to a threat to the strategic nuclear force potential of the Russian Federation” (US Department of State 2009). Then-Secretary of State Hillary Clinton responded that there was no information claiming that the Russians would draft a law on ratification for New START relative to missile defense developments. Clinton noted that these unilateral statements do not change the legal obligations of states parties to the treaty (US Senate 2010). While this unilateral statement about the extraordinary events clause may not have been codified in the Duma, the statement did represent an attempt to undermine irreversibility in the New START treaty by Russia and to hedge against an uncertain future.

New START has come under increasing stress since the treaty was extended by the Biden Administration in 2021. The COVID-19 pandemic caused delays in the on-site inspection regime. This inspection regime allowed for regular communication between American and Russian political and military leadership. Russia has regularly protested that flight crews and on-site inspectors were facing serious difficulties getting the visas required to travel to the United States. Because of these delays, the Russian Foreign Ministry claimed in a statement on 8 August 20228 August 2022 that Moscow resorted to this move in response to Washington’s attempt to hold inspections, “on conditions that do not take into account existing realities and are creating unilateral advantages for the United States” (Bugos 2022). This culminated in a statement by Russian President Vladimir Putin on 21 February 202321 February 2023, that Russia would suspend its participation in New START. At the time of writing, Russia and the United States have ceased on-site inspections but are abiding by the central limits of the treat. It is unclear if New START limits will remain in place in the future or be reversed.

JCPOA

The JCPOA is a nonproliferation agreement negotiated between the Permanent Five members of the UN Security Council (China, France, Russia, the United Kingdom, and United States), Germany, and Iran. The agreement went into force in April of 2015. The JCPOA was pursued with the intention of irreversibility as a guiding principle for all actors. However, the agreement stopped short of full elimination or dismantlement of the capability that they sought to regulate – the enrichment and nuclear fuel cycle of Iran. The agreement placed significant restrictions on several aspects of Iran’s nuclear fuel cycle, such as by limiting what types of centrifuges Iran could possess. Additionally, the agreement converted the Arak reactor to reduce weapons-grade plutonium output. The Iran Deal also placed several international restrictions on Tehran, including adopting the IAEA’s Additional Protocol which allows IAEA inspectors to conduct short-notice inspections and collect environmental samples. In exchange for these restrictions, Iran was promised relief from US, EU, and UN sanctions. While some of these restrictions placed technical irreversible modifications to nuclear facilities, the central provisions of the Iran Deal allowed for enrichment and the use of centrifuges and nuclear reactors under IAEA safeguards. For decades, Republicans and some Democrats have pushed for an agreement with Iran that did not allow for any enrichment, or an agreement with high degrees of irreversibility. This has also been a demand from Israeli leadership such as Prime Minister Benjamin Netanyahu, who repeatedly said that he believes a good deal with Iran means “zero enrichment, zero centrifuges, and zero plutonium” (Lazaroff 2014).

While the agreement technically is still in effect, Iran is no longer abiding by the central limits of the deal and the United States withdrew from the agreement in 2018. The Trump Administration vowed to withdraw from the deal and renegotiate a different agreement. Iranian assistance to Russia during the Ukraine conflict has made any agreement or even negotiations with Iran politically untenable for the foreseeable future. Iran is currently providing Russia with drones and other advanced weaponry. Additionally, Iranian support of terrorist activity and Iran’s continued non-compliance with the Iran Deal’s central limits have made the Biden Administration cautious of diplomatic engagement with Iran. One of Iran’s major demands is that the United States remove the Iranian Revolutionary Guard Corps (IRGC) from the US terrorism blacklist, which the Biden Administration has refused to do.

The Politics of Irreversibility and an IND Research Agenda

Of the six cases explored here, the BWC is the only agreement still in force and operating relatively successfully. The PDMA and INF Treaty are no longer in force, and the JCPOA, New START, and PNIs have precarious futures, if any. These cases tell a dark story about arms control, but demonstrate the challenges for irreversibility beyond arms control, particularly in the face of geopolitical changes and a worsening security environment. Prospects for IND should take these challenges into account and anticipate states’ interests in flexibility within any disarmament agreements and the option to hedge in the event of a change in security environment over time.

Our overarching finding is that many of the non-technical factors in irreversibility have been overlooked and should be incorporated into a future research agenda. Perhaps the most important finding of this article is that irreversibility is, above all, a political exercise. Politics informs whether states enter into agreements, what type of agreements they enter into, and which agreements they stay in. Whether or not states engage with irreversible disarmament, therefore, will depend not just on technical or legal factors, but also be largely driven by political ones and having the flexibility to leave agreements if the security environment changes. While scholarship to date has focused on technical and verification aspects of irreversibility, our approach and the concept of the irreversibility paradox shows that politics- domestic and international- over time are perhaps the most important factors in when, why, and how states participate in arms control and disarmament. The differentiation between technical aspects of agreements, such as fissile material destruction, versus the irreversibility of the agreement itself, such as withdrawal clauses, is an important conceptual point that should be considered in future IND studies.

Other non-technical factors include the value of dialogue and transparency-building measures within arms control and disarmament efforts. While much of the existing scholarship and policy discourse on irreversibility focuses on limiting weapons, less tangible benefits such as dialogues should also be prioritized when designing resilient agreements. The most recent example of this was New START, and there is a very real concern that the loss of New START as a forum for strategic dialogue could increase nuclear risks. Another oft-overlooked aspect of irreversibility is the role of third parties in imposing political costs. In the case of New START, Russia is actively working to undermine the existing rules-based international order, which includes the nuclear order, arms control, and trust – and-confidence-building measures. Beyond the United States and its allies, the international community has been slow to impose political costs on Russia’s reversing of New START.

The case studies point to four important findings. First and foremost, the treaties that proved to be the most enduring over time, even those that may no longer be in effect, have a variety of flexibility mechanisms and prioritize flexibility over irreversibility, typically by relying on withdrawal clauses or end dates to agreements. These mechanisms are often necessary to secure domestic support for agreements, and these political factors of irreversibility prove particularly influential in how states navigate the paradox, as discussed below in greater detail.

Second, the cases confirm that irreversibility is a spectrum rather than a binary – politics, legality, and technical factors work in tandem and cannot be treated in a vacuum. Specifically, the verification mechanisms of the INF Treaty, a technical factor, were essential for President Reagan to put his personal and political weight behind it. Indeed, the origin story of the term “trust but verify” is the intersection of political and technical irreversibility factors. The New START case reinforces the findings of the INF case, particularly about the overlap of political, legal, and technical factors. New START demonstrates the value of arms control beyond the numerical limits and the importance of dialogue and transparency. Future irreversibility efforts, therefore, should not only look at the irreversibility of capabilities, but also the irreversibility of transparency measures as part of arms control and disarmament measures. These agreements might seek some technical provisions that eliminate or disable some weapons production capabilities. However, these agreements stop short of the full elimination of the object that they seek to regulate. Often, these agreements utilize political pressures to ensure that reversal does not occur and offset the lack of full technical procedures for elimination or dismantlement.

Third, the concept of a spectrum of irreversibility would benefit from the addition of a temporal aspect and consideration. While irreversibility might be a spectrum and aggregation of political, legal, and technical factors, it also must take into account how long agreements will last and stand the test of time amidst changes in the security environment. This temporal aspect of irreversibility would seem to be important not only during negotiations, particularly for domestic audiences, but also for how irreversible an agreement proves to be. In an interview for this article, Pavel Podvig notes that the agreements that restricted production of or eliminated fissile material owe their success “to the fact that neither party needs the capability to produce new plutonium. The voluntary moratorium on fissile material production for weapons, maintained by four nuclear weapon states, shows that it is the lack of demand in material rather than a formal agreement, technical measures, or verification procedures that makes irreversibility possible” (Podvig 2023). In these instances, states were willing to forego flexibility because they did not anticipate a need to produce new plutonium and observed an increasingly stable geopolitical climate. But as the climate changes over time, states may re-evaluate participation in agreements. The durability of arms control, and thereby its irreversibility, could decline with time as the geopolitical and technical landscapes changes. For over two decades, the INF Treaty worked extremely well and was a prime example of irreversible arms control. It should not be seen as a failure, but rather a product of its time. The reductions of INF were ultimately reversed not necessarily because of a weakness in the treaty’s design – its legal or technical factors – but rather because of a changing geopolitical landscape and Russian nuclear ambitions.

Nonetheless, even agreements with a short duration can contribute to irreversible disarmament. In an interview for this study, Tong Zhao noted, states tend:

to think that arms control agreements that can only limit or prohibit certain capabilities within a finite period may still be worth pursuing, as long as they contribute to limitation of arms competition or to reduction of military incidents or inadvertent escalation for the foreseeable future. They seem to understand that situations evolve, and that their own — as well as their rival’s — interests change. Therefore, arms control arrangements have value even if they do not introduce permanent limits or prohibition (Zhao 2023).

This temporal aspect and how states slide along the irreversibility spectrum over time is one topic particularly worthy of further study. But ultimately, temporal aspects are an important political consideration for states when entering into agreements, which is not necessarily reflected in current IND scholarship.

And fourth, domestic politics heavily shape how states reconcile the irreversibility paradox, and whether or not they stay in arms control and disarmament agreements. This domestic aspect is missing from many studies of irreversibility and the concept of irreversibility as a spectrum. The withdrawal clause – essentially a reversible option — was essential for the conclusion of the INF Treaty. Senator Sam Nunn, then chairman of the Senate Armed Services Committee, urged President Reagan to insist on an extraordinary events clause, fearing that Soviet advantages in tanks and other mobile weapons could provide Russia with a battlefield advantage that would require US nuclear deterrence (Wilson 1987). Indeed, intentional reversibility proves to be an important component of successfully completed arms control agreements, such as INF. Additionally, the case of the JCPOA also demonstrates the influence of domestic politics on the endurance of agreements.

The cases explored here offer important implications for nuclear policy and diplomacy, particularly for the NPT and for any future IND initiatives. Questions of irreversibility are likely to feature prominently in the 2026 NPT Review Cycle, and in other disarmament forums, such as the Conference on Disarmament, the Creating an Environment for Nuclear Disarmament (CEND) initiative, and in non-nuclear arms control, including the BWC and Chemical Weapons Convention. The topic of IND has gained attention from nuclear possessors and non-possessors alike. Some states view the IND initiative as a bridge-building exercise between NWS and NNWS (Gwadera 2022). Some NWS see the IND initiative as a way to demonstrate commitment toward fulfilling their Article VI obligations to pursue negotiations in good faith toward the cessation of an arms race and general and complete disarmament. In short, IND discussions are likely to be policy relevant within the next NPT review cycle.

These findings point to important policy implications and recommendations for the future of the IND initiative, whether led by the UK and Norway, or as a multilateral effort among a wider set of nuclear and non-nuclear possessors, government, and non-governmental actors. First and foremost, agreements with greater flexibility prove more enduring over time. This is because agreements that allow for flexibility are more politically acceptable to states and may garner wider support, particularly in states with domestic political polarization and contestation of arms control. But additionally, agreements with higher flexibility are more enduring because they allow states to adapt to geopolitical uncertainty and changes. The enemy of irreversibility is time. This should inject a degree of humility into any irreversibly efforts and acknowledgment that if/when states engage in arms control or disarmament, they will require an “opt-out”.

This should not be cause for pessimism; rather, an examination of irreversibility in arms control points to important pathways for IND initiatives and research. A deeper exploration of historical cases by civil society and governments would be an important bridge-building activity to explore the spectrum of political, legal, and technical factors in the context of disarmament. This initiative would help to identify ways to lay the groundwork for future disarmament by identifying potentially controversial questions and solutions, such as whether intrusive verification would be required for IND. Further work could also look at the intersection of deterrence for security purposes and irreversibility. This finding has policy implications not only for disarmament, but also for future arms control negotiators.

Conclusion

Arms control and disarmament agreements require the acceptance of imperfection. Flexible agreements are likely to be more acceptable to states but may come at the cost of technical irreversibility and transparency. In a time of geopolitical turmoil, it will be difficult for the United States to tie its hands with a legally binding treaty with strict and irreversible technical, legal, and political commitments. The recent Biden Administration self-declared moratorium on kinetic anti-satellite testing is one such example of how low irreversibility arms control can help set rules of the road for strategic competition without hand-tying. In an interview, Tong Zhao observed, “Many states appear to think that arms control agreements that can only limit or prohibit certain capabilities within a finite period may still be worth pursuing … . They seem to understand that situations evolve, and their own (as well as their rival’s) interests change. Therefore, arms control arrangements have value even if they do not introduce permanent limits or prohibition” (Zhao 2023). This requires humility not only in the depth of irreversibility, but also their duration.

This paper has unpacked how concepts of irreversibility have manifested in arms control agreements that have sought disarmament, nonproliferation, and numerical constraints on nuclear weapons. Most of these agreements ultimately collapsed because of the interface between political, technical, and legal challenges over time. More irreversibility in arms control agreements means that rebuilding a nuclear weapons program would take more time and incur more costs to the defector. This type of agreement, which seeks to place more technical limits on states parties, will likely face uphill battles in the domestic audiences of participating states. Less irreversibility may seek to place some technical limits and seek political ramifications for defection. This style of agreement may capture some of the disarmament benefits of destroying key facilities or materials but will fall short of the full destruction of the full scope of the weapons and materials that the agreement seeks to govern and constrain.

On the opposite end of the spectrum, flexible agreements may seek to shape norms and behaviors through moratoriums or abiding by limits of past treaties after their expiration dates, and allow states to respond to the security environment. These types of agreements are highly reversible in nature and may be easier to sell to domestic audiences. In some cases, these agreements may not even require legislative buy-in, as in the case of the 2022 United States Kinetic Anti-Satellite Testing Moratorium. The BWC, which is highly reversible and contains no real verification mechanisms, may only be possible with a high degree of reversibility. The dual-use nature of biological weapons may make achieving irreversibility in this domain impossible. These agents exist in nature. For agreements with more irreversibility, discussions should focus more on transparency, verification, and consistency of activities rather than on material balances and facilities capable of producing weapons materials.

There are challenges with this approach, which will require further research. While states may prioritize flexibility in agreements with low irreversibility for the sake of national interest, in other cases they may prioritize intrusive verification mechanisms to ensure adversary compliance. The sufficiency of verification was an important factor in the 1999 debate over the Comprehensive Test Ban Treaty, for example, along with the New START Senate debate. This will depend largely depend on geopolitics and multilateral or bilateral relationships, as was the case with INF and New START; but it may also depend on the nature of nuclear risk, and crises can prompt more action, as was the case in INF. Another challenge with this approach is that agreements with less irreversibility may not be seen as meaningful progress towards disarmament commitments, most notably Article VI of the NPT. They may not qualify as what Nick Ritchie refers to as a deep devaluing of nuclear weapons. Ultimately, this too will depend on political context and dynamics within the NPT itself.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

The work was supported by the King’s College London .

Notes on contributors

Joseph Rodgers

Joseph Rodgers is an Associate Director and Associate Fellow with the Project on Nuclear Issues in the International Security Program at the Center for Strategic and International Studies (CSIS). He is also a PhD student in the biodefense program at George Mason University. Previously, he worked as a graduate research assistant at the James Martin Center for Nonproliferation Studies and interned with the United Nations Institute for Disarmament Research. Joseph holds an MA in nonproliferation and terrorism from the Middlebury Institute for International Studies.

Heather Williams

Heather Williams is the Director of the Project on Nuclear Issues and a senior fellow in the International Security Program at the Center for Strategic and International Studies (CSIS). She is also an associate fellow with the Project on Managing the Atom in the Belfer Center for Science and International Affairs at the Harvard Kennedy School. Before joining CSIS she was a visiting fellow with the Project on Managing the Atom and a Stanton Nuclear Security fellow in the Security Studies Program at MIT. Until 2022, she was a senior lecturer (associate professor) in defense studies at King’s College London and taught on arms control, deterrence, and disarmament. Dr. Williams has a PhD in war studies from King’s College London, an MA in security policy studies from the George Washington University, and a BA in international relations and Russian studies from Boston University.

Notes

¹ We are grateful to the King’s College London workshop discussion for highlighting this as a potential area for future research. Two such comparisons that would be of use for future research would be comparing the NPT and CTBT, and comparing START and New START.

² For a detailed history of the Soviet biological weapons program, see Leitenberg and Zilinskas (2012).

³ December, 1987. Article XIII, Treaty Between the United States of America and the Union of Soviet Socialist Republics on the Elimination of Their Intermediate-Range And Shorter-Range Missiles (INF Treaty), US Department of State, Bureau of Arms Control, Verification, and Compliance.

⁴ Ibid.

⁵ December, 1897. INF Treaty Signing, CSPAN.

⁶ See, for example, Koch (2012).