The last atomic Waltz: China’s nuclear expansion and the persisting relevance of the theory of the nuclear revolution

ABSTRACT

China’s nuclear expansion has fueled debates about its nuclear strategy. In this article, I argue that there is still little evidence to suggest that China’s strategy has changed. The expansion can best be explained as an effort to bolster China’s second-strike capability, and it is primarily driven by increasing concerns about US missile defense as well as nuclear and conventional counterforce weapons. This is in line with assumptions of the nuclear revolution theory, which stresses the importance of secure second-strike forces. So far, China’s nuclear modernization does not align with the “delicate nuclear balance” school, which sees strong incentives for states to adopt competitive nuclear policies and attempt to obtain strategic superiority. The findings further highlight how US policy choices will have a significant influence on China’s future nuclear posture. A strong US response to China’s expansion will feed arms race dynamics and increase the risks of China fielding destabilizing weapons.

KEYWORDS:

• China
• nuclear weapons
• nuclear revolution theory
• deterrence

A nuclear counterrevolution is brewing. The theory of the nuclear revolution—which has dominated scholarship on nuclear weapons for decades—is under fire, with critics highlighting that its predictions have not come to pass. According to nuclear-revolution scholars, states with a second-strike capability are essentially secure, which should greatly dampen the security dilemma among nuclear powers (Jervis, 1989; Waltz, 1990). Yet, as critics point out, the superpowers continued to engage in intense security competition throughout the Cold War (Rovner, 2018; Tannenwald, 2020; Walt, 2010). Moreover, a group of scholars have taken aim at the foundational claim of the nuclear-revolution theory, namely that nuclear weapons produce stalemate. Based on a reexamination of Cold War history and an analysis of recent technological shifts, these scholars argue that the balance of terror was and remains delicate. The incentives to develop large and diverse nuclear arsenals—whether to gain superiority or to secure coercive leverage—are therefore significant (Green, 2020; Lieber & Press, 2020).

China’s ongoing nuclear expansion has further fueled the debate about the nuclear revolution. With a modest arsenal and leaders that have viewed nuclear weapons as useful only for deterrence or countering nuclear coercion, China has often been seen as a revolutionary per excellence (Christensen, 2012; Cunningham & Fravel, 2015; Fravel & Medeiros, 2010; Goldstein, 2000). However, recent revelations demonstrate that China is building more than 300 missile silos in three different fields. The US Department of Defense (DOD) (2023, p. 111) now assesses that China will possess over 1000 warheads by 2030, a fourfold increase in a decade. Moreover, in 2016, China fielded the DF-26 missile, which provides it with a theater-range nuclear precision-strike capability. As a result, critics of the nuclear revolution theory claim China could adopt nuclear warfighting concepts (Green, 2021, p. 32; see also Johnson, 2019). They further argue that “it is not a foregone conclusion that China will be a good competitor in seeking and maintaining mutual second-strike capability” (Long, 2020, p. 44). US officials voice similar concerns and allege that China may be building towards a “de facto first-strike capability” (Weisgerber, 2021).

China’s recent expansion and modernization efforts therefore raise crucial questions. Is China’s nuclear strategy changing? And are Chinese leaders adopting a nuclear strategy that breaks with the prescriptions of the nuclear revolution theory? While acknowledging that China’s future nuclear trajectory is more uncertain, I answer both questions in the negative. To assess the drivers of China’s nuclear expansion, I have examined the writings of civilian arms control and nuclear experts and scholars and officers associated with the People’s Liberation Army (PLA), as well as recent official statements by Chinese leaders. Although the changes in China’s posture may enable later shifts in nuclear strategy, I find little evidence that its strategy has changed, or that Chinese leaders are fundamentally rethinking the purposes nuclear weapons serve in their defense policy.

More in particular, I show that the expansion and modernization still best can be explained as an effort to bolster China’s second-strike capability, informed by increasing worries about US missile defense, as well as conventional and nuclear counterforce capabilities. Although factors such as great power status and a perceived need for a more robust nuclear “shield” to enable offensive operations against Taiwan may also have influenced China’s nuclear expansion, I argue that concerns about the robustness of its second-strike forces are the primary driver. Increasing fears of US nuclear first use in a Taiwan conflict has led to some debate in China about the need for non-strategic nuclear weapons as proportional response options, but there is little evidence to suggest that China is searching for coercive nuclear tools. The evidence of China seeking strategic nuclear superiority and a first-strike capability is weaker still.

I further argue that the expansion arguably makes China a more “mainstream” nuclear revolutionary. Nuclear revolution theory comes in different stripes. A minority of scholars, including Kenneth Waltz (1981; 1990), saw a small and potentially vulnerable nuclear arsenal as sufficient for deterrence—a position often labeled as “minimum deterrence” (see Lewis, 2008, pp. 38–41). Most other scholars within this tradition have regarded mutually assured destruction (MAD) as the key benchmark, and do not reject that states may need a small suite of limited nuclear options, such as nuclear weapons with a lower yield. Examining the evolution of China’s nuclear posture, I highlight how the arsenal during the Cold War and into at least the early 2000s was consistent with the minimum deterrence position. The recent expansion makes China’s posture more aligned with the MAD position, but it does not match the competitive nuclear policies prescribed by delicate nuclear balance scholars.

Demonstrating that China’s policy conforms to its expectations does not salvage the nuclear revolution theory from criticism leveled against it. Moreover, critics may still claim that nuclear revolution theorists are too confident in the pacifying effects of MAD, and that a deeper nuclear stalemate could make it safer for Chinese leaders to engage in conventional aggression without fear of strategic nuclear escalation (Denmark & Talmadge, 2021). Nevertheless, China’s nuclear policies illustrate the limits of some of the criticism. Furthermore, displaying the similarities between Chinese nuclear policy and that prescribed by mainstream nuclear-revolution theorists may provide insight into China’s future direction. Such insight is crucial. Because great power rivalry between the United States and China appears set to dominate the coming decades, it is vital to understand whether China will adopt a more competitive and potentially risk-inducing nuclear policy or maintain a more restrained approach.

The remainder of this article proceeds as follows. First, I briefly outline the recent debate about the nuclear revolution theory. In the second part, I delineate the three different ideal-type positions to the question of what nuclear posture a state needs for credible deterrence. In the third part, I outline the evolution of China’s nuclear posture from 1964 to the mid-2010s. In the fourth section, I debate China’s recent nuclear modernization efforts, and examine debates in China about its nuclear security climate. In the conclusion, I outline the implications my findings have for debates about the nuclear revolution theory, and for discussions about China’s future nuclear trajectory. I highlight how China’s nuclear posture will be influenced by policy choices made in the US, and how a US buildup in response to China’s expansion will further fuel arms race dynamics.

The nuclear revolution and its critics

Nuclear revolution theorists claim the bomb has fundamentally altered the nature of warfare. Bernard Brodie (1959), the dean of nuclear strategy, wrote, “People often speak of atomic explosives as the most portentous military invention ‘since gunpowder’. But such a comparison inflates even so epoch-making an event as the introduction of gunpowder” (p. 147). Brodie and other theorists such as Thomas Schelling (1966), Glenn Snyder (1961), Kenneth Waltz (1981), and Robert Jervis (1989) highlighted how nuclear weapons can wreak immense destruction and are impossible to defend against, and therefore produce stalemate between nuclear-armed states. Consequently, the bomb has transformed how states can provide security for themselves, with deterrence replacing defense. According to Schelling (1966), “Deterrence rests today on the threat of pain and extinction, not just on the threat of military defeat” (p. 22).

Based upon these observations, the second component of the theory is that the bomb has altered security relations among states. With nuclear weapons ensuring their possessors’ security, the world should be more peaceful. Waltz (1988) argued that “the probability of major war among states having nuclear weapons approaches zero” (p. 627). In a world of nuclear deterrence and mutual vulnerability, security dilemmas are therefore largely eliminated, and states can refrain from behaviors such as arms racing or forming alliances. Because neither nuclear nor conventional inferiority impinges on states’ fundamental security, concerns over relative gains are greatly reduced. Jervis (1989) summarized the expectations of theory:

If nuclear weapons have had the influence that the nuclear-revolution theory indicates they should have, then there will be peace between the superpowers, crises will be rare, neither side will be eager to press bargaining advantages to the limit, the status quo will be relatively easy to maintain, and political outcomes will not be closely related to either the nuclear or the conventional balance (p. 45).

Although contested during the Cold War already, a series of recent scholarship has further challenged the theory of the nuclear revolution. The criticism is twofold. First, while recognizing that the key prediction of peace between great powers still holds, scholars have pointed to the gap between predicted behavior and the actual conduct of nuclear-armed states. Far from ignoring conventional and nuclear balances, accepting the status quo, and refraining from pressing bargaining advantages, nuclear powers continued to engage in highly competitive behavior (Rovner, 2018; Tannenwald, 2020; Walt, 2010). For example, scholars claim Russia’s war in Ukraine and its use of nuclear threats is at odds with nuclear revolution theory (Arceneaux, 2023).

Certainly, nuclear-revolution theorists have long been aware of gaps between their expectations and actual policy, explaining it as irrationality among decision-makers and failure to grasp the bomb’s revolutionary implications (Jervis, 1984). However, in a second and potentially more profound line of criticism, a group of scholars have taken aim at the technological claims at the heart of the theory. Mirroring the views of Cold War scholars such as Wohlstetter (1958), these scholars have argued that the balance of terror was and remains far more “delicate” than the nuclear-revolution theory asserts. Based on a detailed review of Cold War nuclear history, Green (2020) has demonstrated that US policymakers regarded the balance during the 1970s as tenuous and sought to pursue nuclear superiority (see also Long & Green, 2015). Moreover, despite the alleged impossibility of escaping mutual vulnerability, Soviet leaders worried a great deal more about the vulnerability of their arsenal than the theory would lead us to expect, even during the latter decades of the Cold War (Green & Long, 2017b). Relatedly, Lieber and Press (2017) have suggested that technological shifts during the last decades has made nuclear arsenals much more vulnerable than in the past. According to these scholars, MAD is therefore not “a fact,” as Robert Jervis (1989, p. 74) asserted, but “a variable” (Long, 2020, p. 31). This line of criticism has profound prescriptive implications, with delicate nuclear balance scholars arguing that nuclear-armed states have strong incentives to compete, and to pursue large and diverse nuclear arsenals.

Even if all scholars associated with the nuclear revolution theory agree that the bomb has had an immense impact on international relations, their views of the threshold for stalemate and the military requirements of deterrence differ significantly. In the next section, I outline the minimum deterrence position and the more mainstream MAD position, and how they differ from each other, as well as from the delicate nuclear balance school.

Nuclear posture and the military requirements for deterrence

When assessing the nuclear posture a state requires, nuclear theorists frequently ask three key and related questions. First, and most basically, how large and sophisticated forces does a state need to produce nuclear stalemate? Second, under what conditions should states invest in counterforce weapons and potentially attempt to pursue or maintain a first-strike or damage-limitation capability—and thus break out of stalemate? Third, do states need flexible, limited nuclear options—typically weapons with lower-yield, or so-called nonstrategic (tactical) nuclear weapons—for deterrence, coercion, or warfighting?

This section outlines three positions: (i) Minimum deterrence; (ii) MAD; and (iii) Delicate nuclear balance and the key characteristics of their associated nuclear postures. Certainly, scholars within each camp may hold divergent views on some issues, and some of the scholars and analysts cited have not always been fully consistent in their writings. Nevertheless, these ideal types represent three different approaches reflected in the literature about what sort of nuclear posture a state needs to deter.

Minimum deterrence

Minimum deterrence holds that even a small number of nuclear weapons can ensure robust deterrence (Lewis, 2008). According to this position, the devastating consequences of a nuclear counterattack will instill fear in any adversary, and thus produce strong deterrent effects even in situations of great asymmetry in arsenal size. The revolutionary effects of nuclear weapons apply before MAD, and states do not necessarily even need a secure second-strike capability. For example, according to Kenneth Waltz (1981), arsenal size, configuration and readiness levels matter little, as “an attacker is deterred even if he believes only that the attacked may retaliate” (p. 18). Placing the bar only slightly higher, Bernard Brodie (1959) argued that if a “small nation could threaten the Soviet Union with only a single thermonuclear bomb, which, however, it could and would certainly deliver on Moscow,” it would deter the Soviets (p. 275).

Minimum deterrence scholars stress the role of uncertainty and believe the mere possibility of devastating retaliation will stay the hand of opponents. For example, McGeorge Bundy claimed that “uncertainty about what could happen” produced a strong “existential deterrence” effect in every major crisis between the United States and Russia during the Cold War (1983; see also Waltz, 1981, p. 16). States can therefore base their deterrence strategies on first-strike uncertainty, which means “doubt that even the best planned surprise attack will neutralize the victim’s ability to launch an unacceptably punishing retaliatory strike” (Goldstein, 2000, p. 44; see also Hagerty, 1995; Wu, 2013). According to this standard, an arsenal with limited survivability may be sufficient for deterrence, as it will instill fear even in a much more powerful opponent.

Because of the belief that even small and potentially vulnerable arsenals will provide robust nuclear deterrence, scholars in the minimum deterrence camp are dismissive about attempting to escape nuclear stalemate, and thus also about pursuing counterforce capabilities. Pursuit of such capabilities is meaningless and wasteful against any nuclear adversary, since states “can deploy their forces in ways that preclude preemption,” without being “rigged for hair-trigger response” (Waltz, 2013, p. 20). Moreover, given that even the prospect of a small number of nuclear weapons surviving a first strike is enough to deter, counterforce capabilities do not provide states with any coercive leverage.

The minimum deterrence position further rejects the need for limited nuclear options, much less arsenals dedicated to nuclear warfighting, as they see nuclear weapons as useless for any purposes beyond deterrence. This position is associated with a strong disbelief in the prospects of controling nuclear escalation, or even to make threats of escalating that an adversary will find credible (see Cunningham & Fravel, 2019, p. 72). Consequently, limited nuclear options have no utility: Any threat of limited first-use would not be credible, with the possibility of escalation to all-out nuclear war looming large. Given the lack of credibility of such threats, states also do not need to develop any limited nuclear options to deter others from making them, or to be able to retaliate “tit for tat.”

In sum, the minimum deterrence position prescribes developing a small arsenal, keeping forces at low alert levels, abstaining from developing counterforce capabilities, and eschewing limited nuclear options at both the theater and strategic level. States with a minimum deterrence posture can make do with quite rudimentary command and control capabilities and no early warning capabilities.

MAD

In contrast to their more radical peers, most nuclear revolution scholars stress that stable deterrence requires two states possessing an assured destruction capability (Glaser, 1990, pp. 52–53; Jervis, 1989; Van Evera, 1999). According to this position, the nuclear revolution only truly came to pass when the Soviet Union developed secure second-strike forces during the early 1960s. For example, Robert Jervis (1989, p. 9) argued that it was the advent of “mutual second-strike capability and not nuclear weapons per se” that revolutionized world affairs.

A small nuclear arsenal is therefore not necessarily enough, as credible deterrence rests on certainty of retaliation, not uncertainty. To ensure credible deterrence, a state needs to develop deliverable weapons that can not only survive any attack by an adversary, but also later penetrate its defensive systems. That may require building a large arsenal, as well as other measures that render forces invulnerable, such as obtaining very quiet nuclear-propelled ballistic missile submarines (SSBNs). While acknowledging potential risks to strategic stability, scholars in this camp highlight that establishing a secure second-strike capability may require that states with vulnerable nuclear forces maintain high levels of readiness or even adopt policies such as “launch on warning” (Jervis, 1989, pp. 144, 146; see also Glaser & Fetter, 2017, p. 207). In contrast with delicate nuclear balance scholars, mainstream nuclear revolution scholars see no benefits of attempting to obtain strategic superiority. Once established, they believe secure second-strike forces are robust and difficult to undermine (e.g., Jervis, 1984; Van Evera, 1999).

There are subtle, but important differences between MAD scholars and their more radical peers also when it comes to the viability of pursuing counterforce capabilities for damage limitation (or even as a first-strike option). At the outset, scholars in this camp are also dismissive about damage limitation, or the ability to “meaningfully achieve a better outcome in an all-out nuclear war” and have argued against US pursuit of such capabilities even against secondary nuclear states, such as China (Glaser & Fetter, 2016, p. 54). They have also been frequent critics of efforts to develop missile defenses. Nevertheless, they do not fully reject the need for maintaining—and in a very narrow set of circumstances even employing—counterforce options vis-à-vis very weak nuclear adversaries such as North Korea (Glaser & Fetter, 2005, p. 89).

Finally, scholars in this camp are skeptical about limited nuclear options, but do not outright dismiss their utility in a narrow set of circumstances. At the outset, they are highly doubtful about the prospects for controlling nuclear escalation and therefore see no need to develop nuclear arsenals for warfighting. Nevertheless, building on Schelling’s (1966, pp. 92–125) concept of “risk manipulation,” they regard it as possible—if exceedingly risky—to gain a coercive advantage by threatening to conduct limited nuclear strikes. Furthermore, even if nuclear crises are competitions in risk-taking and determined primarily by the balance of resolve, not the balance of capabilities, some scholars in this camp see a small set of limited nuclear options as necessary to deter threats of limited nuclear use from others (see Cunningham & Fravel, 2019, p. 72).

In sum, according to the MAD position, states may—depending on the capabilities of their adversaries—require a relatively sophisticated and medium- or even large-sized nuclear force. If the second-strike capability is secure, a state does not need to build a force that matches or exceeds the arsenal size of its adversaries. To ensure survivability, a state may have to keep parts of its forces at elevated alert levels, and is likely to develop strategic intelligence, surveillance, and reconnaissance capabilities for early warning. It may also pursue a limited set of counterforce capabilities (but only against weaker adversaries) and some limited nuclear options, but will refrain developing an extensive and diverse set of either.

Delicate nuclear balance

In contrast with both strands of the nuclear revolution theory, scholars in the delicate nuclear balance school see strong incentives for states to engage in competitive nuclear policies and attempt to obtain strategic superiority. Although they agree that MAD is a key threshold for stable deterrence, they see it as hard to both reach and stay beyond this threshold. The challenge of reaching stalemate, they argue, springs from the technical nuclear balance being delicate, but also from the perceptions of policymakers, who are likely to be concerned about whether stalemate will persist (Green, 2020, pp. 28–47). Moreover, they claim states need a large arsenal to maintain not only a peacetime second-strike capability, but also, ideally, “resilient survivability” in wartime, particularly if they want to pursue coercive nuclear escalation strategies (Lieber & Press, 2020, pp. 105–106).

In marked contrast with nuclear revolution theorists, delicate nuclear balance scholars are supportive of pursuing damage limitation and developing a broad suite of counterforce weapons, including ballistic missile defense systems. For example, several scholars belonging to this camp have argued forcefully that the United States should attempt to maintain such a capability against China (Green & Long, 2017a; Kroenig, 2017). In addition to limiting damage in the event of an all-out war, delicate nuclear balance scholars have pointed to other benefits of counterforce capabilities and a large arsenal, arguing that nuclear superiority helps “win” crises (Kroenig, 2018). Beyond crisis advantages, Green (2020, pp. 50–52) has claimed that nuclear competition may yield important peacetime political benefits, including enhancing general deterrence, diverting enemy resources, forcing grand strategic shifts, and bolstering alliance cohesion.

Current proponents of the delicate nuclear balance position also see a strong need for robust limited nuclear options and a flexible arsenal at both the theater and strategic level (Lieber & Press, 2009). Compared to the nuclear revolution theorists, scholars within this camp are more sanguine about the prospects of controlling nuclear escalation and fighting limited nuclear wars (Kartchner & Gerson, 2014). Relatedly, they highlight that without limited nuclear options at hand, adversaries may believe they can successfully fight and win nuclear wars, and that this perceived advantage may provide them with bargaining leverage (Green, 2020, p. 45). Therefore, delicate nuclear balance scholars see the development of a diverse and potentially large set of limited nuclear options that can shape battlefield outcomes as beneficial. Such an arsenal, they claim, “would permit planners to devise a coercive campaign with several rungs in the escalatory ladder” (Lieber & Press, 2020, p. 104).

In short, the delicate nuclear balance scholars claim states need large nuclear forces that go beyond the requirements of assured destruction in peacetime. If they want to acquire a damage-limitation or even a first-strike capability, they need an arsenal that significantly exceeds its rivals in terms of both numbers and quality, and that includes robust counterforce options such as precise nuclear weapons and missile defense systems. To engage in coercive nuclear escalation and warfighting, states further need a range of limited nuclear options, including both theater- and longer-range systems, and warheads of various yields. To support coercive nuclear escalation and counterforce campaigns, states also need a range of supporting capabilities, such as redundant command and control systems and advanced strategic intelligence, surveillance, and reconnaissance capabilities to identify and track the nuclear forces of adversaries, and it is likely to keep a substantial part of its force at elevated alert levels.

The radical phase (1964–2010s): Chinese force posture and beliefs about the requirements for deterrence

Of the three positions outlined above, which one best describes China’s posture and Chinese leaders’ beliefs about nuclear weapons? After conducting its first nuclear test in 1964, China built a small and unsophisticated nuclear arsenal, and its nuclear modernization was slow. Although it may have aspired for an assured retaliation or second-strike capability, its nuclear forces were for decades vulnerable, and Chinese leaders appear to have believed that such an arsenal sufficed for deterrence. Although not a term Chinese leaders have used to describe their own strategy (for a recent exception, see PRC Ministry of Foreign Affairs, 2020), China’s posture has aligned quite closely with the minimum deterrence position.

Throughout the Cold War, China had a small nuclear arsenal. In 1975, almost a decade after its first nuclear test, analysts have estimated that China possessed around 75 nuclear warheads, a small fraction of its Cold War superpower adversaries (Norris et al., 1994, p. 359). Moreover, China’s delivery vehicles were few, unsophisticated, and vulnerable. The silo-based DF-5, which was the first missile capable of reaching the continental United States, was first deployed in 1981, more than 15 years after China’s first nuclear test. By 1991, China only had four operational missiles deployed in silos (Lewis & Hua, 1992, pp. 19, 24). Throughout the Cold War and beyond, China did not develop an operational nuclear triad. Although China started its SSBN program during the late 1950s, China’s first submarine—which was launched in 1985—was riddled with problems, and never conducted a deterrent patrol (Kristensen et al., 2006, p. 89). China’s bomber force consisted mainly of H-6 bombers, which had limited range and were highly vulnerable to Soviet air defense (Kristensen et al., 2006, pp. 93–94).

China’s limited nuclear arsenal was accompanied by a restrained nuclear doctrine. Immediately after its first nuclear test, China stated that it would not use nuclear weapons first in a conflict. For several decades, China refrained from developing a detailed operational doctrine to help offset its inferiority vis-à-vis the superpowers (Fravel & Medeiros, 2010, p. 48). Moreover, it kept nuclear weapons at a low level of alert, with warheads reportedly not mated to delivery vehicles in normal circumstances and stored separately (Stokes, 2010), and did not develop early warning systems. Although resource constraints may initially have inhibited China’s efforts to build a more sophisticated arsenal, the limited pace of its build-up—which consistently defied the expectations of US intelligence services—reflected a chosen policy, and the beliefs of its leaders that a small number of nuclear weapons are sufficient for deterrence (Fravel & Medeiros, 2010; see also Lewis, 2007).

Scholars disagree about exactly where Chinese leaders have set the bar for credible deterrence. Fravel and Medeiros (2010, p. 51) claim that Chinese leaders aspired for an assured retaliation capability, but that Mao Zedong and Deng Xiaoping did not possess “views on the operational requirements of credible deterrence vis-à-vis China’s potential adversaries.” Moreover, they stress that organizational and political constraints inhibited the PLA from developing an operational doctrine (pp. 51–52). By contrast, Wu Riqiang (2013) claims that Chinese leaders have viewed first-strike uncertainty as sufficient, and that their “criteria for nuclear deterrence [is] uncertain retaliation, rather than assured retaliation” (p. 591). Despite this disagreement, recent scholarship demonstrates that China for most of its nuclear history has not possessed a secure second-strike capability. Modeling nuclear exchanges with the Soviet Union in 1984 and the United States in 2000 and 2010, Wu (2020, see also Tecott & Halterman, 2021, pp. 74–78) finds that particularly in 2000, before China fielded the road-mobile, solid-propellant DF-31 ICBMs, its arsenal was highly vulnerable to a first strike. Even in 2010, after the deployment of the DF-31s, Wu found that China’s probability of successful retaliation with a single warhead was far from assured.

As a corollary of its leaders’ beliefs of the limited utility of nuclear weapons, China has never deployed limited nuclear options, and Chinese strategists have been very skeptical about the prospects of controlling nuclear escalation. As Cunningham and Fravel (2019) have highlighted, they have therefore also believed it highly unlikely that other actors will conduct limited nuclear strikes. China did develop and test an enhanced radiation-weapon during the 1980s—which would, as Jonathan Ray (2015) has argued, have been “ideal tactical and antipersonnel weapons” (p. 6)—but it never deployed it.

Although China’s nuclear posture was under debate at various stages, voices calling for a shift in China’s nuclear strategy did not manage to win the favor of the leadership. For example, during the late 1980s, several People’s Liberation Army scholars and officers argued that China should adopt what they labeled a “limited-deterrence” doctrine—that is, a limited flexible-response posture tailored towards nuclear warfighting. As part of this doctrinal shift, they argued in favor of developing a range of capabilities, including tactical nuclear weapons, space-based early warning, and a limited counterforce capability (Johnston, 1995). There were also internal debates among academics and military commentators during the early 2000s about whether China should abandon its no-first-use policy and consider developing nuclear weapons with a tactical role, as well as about the size and composition of China’s nuclear arsenal more broadly (Horsburgh, 2015, pp. 127–129). However, at both junctures, China maintained its no-first-use policy (albeit surrounded in some ambiguities), continued to develop a limited nuclear arsenal, and refrained from developing tactical nuclear weapons.

In sum, at least until the mid-2000s, China’s nuclear posture—and the beliefs of its leaders—aligned very well with the minimum deterrence position, and not with the mainstream renditions of the nuclear revolution theory. China was for decades willing to live with a small, vulnerable arsenal with no limited nuclear options or counterforce options. Only with the deployment of road-mobile, solid-propellant ICBMs in the late 2000s did China approach a second-strike capability—and even then, doubts about the survivability of its arsenal remained.

Becoming a mainstream revolutionary (late 2010s–present): China’s nuclear expansion

In recent years, China’s nuclear posture has undergone major changes. First and foremost, China has engaged in a quantitative expansion, and raised the alert level of its nuclear forces. In the summer and autumn of 2021, researchers revealed that China was building several hundred new missile silos in three different fields (Korda & Kristensen, 2021; Lee, 2021; Lewis & Eveleth, 2021). Although it is still unclear whether China will deploy missiles in all of them, and how many warheads each missile will carry, there is little doubt that the silo-building represents a major expansion of China’s arsenal. The US DOD (2023) recently stated that China has loaded ICBMs into at least some silos and assessed that China “will have over 1,000 operational nuclear warheads by 2030” (p. 104). It further stated that China has continued its “rapid nuclear expansion” and estimated that its “stockpile had more than 500 operational nuclear warheads as of May 2023” (US DOD, 2023, p. 104).

The silo buildup is happening on top of ongoing nuclear modernization and efforts to develop an operational nuclear triad. China is bolstering its sea-based nuclear forces, and has “probably” fielded the JL-3 missile, which can reach the continental United States from Chinese littoral waters, on its JIN-class SSBNs (US DOD, 2023, p. 108). Chinese leaders have reportedly given the PLA Air Force a nuclear mission, and the air force recently fielded the H-6N nuclear-capable bomber and is developing a long-range stealth bomber. China is also developing other new advanced delivery systems, and tested an ICBM-range hypersonic glide vehicle in 2021. The test also likely demonstrated China’s ability to field a so-called fractional orbital bombardment system—a system that places weapons in orbit before striking targets. Moreover, the US DOD is reporting that both conventional and nuclear PLA Rocket Force brigades are conducting “combat readiness duty” and “high alert duty” and allege that China is adopting a “launch-on-warning” posture for at least portions of its forces (US DOD, 2023, pp. 106, 112).

US officials and analysts have indicated that China’s silo building and development of capabilities such as a fractional orbital bombardment system may constitute a pursuit of counterforce weapons that could put the US second-strike capability at risk. Air Force Secretary Frank Kendall, for instance, has stated that if China continues to “substantially increase their ICBM force,” it will acquire a “de facto first-strike capability” (Weisgerber, 2021). Other military leaders have argued that China’s nuclear expansion represents a “strategic breakout,” which will provide China with “the capability to execute any plausible nuclear employment strategy” (US Strategic Command, 2021). Prominent analysts have also warned of Chinese first-strike intentions and argue that capabilities such as the fractional orbital bombardment system “poses a potential decapitating threat to the US nuclear command and control system” (Center for Global Security Research, 2023, p. 14).

In addition, with the development of the dual-capable DF-26 missile, China for the first time possesses a theater-range nuclear precision-strike capability. The US DOD (2023) have further claimed that China probably seeks to develop lower-yield weapons “to provide proportional response options that its higher-yield weapons cannot deliver” (p. 111) and believe the DF-26 is the most likely delivery vehicle for such warheads. Analysts speculate that China may rely on particularly the DF-26 to adopt a limited first-use posture for coercion (Montgomery & Yoshihara, 2022). Both the 2018 and the 2022 Nuclear Posture Review raised similar concerns (US DOD, 2018, p. 32; US DOD, 2022a, p. 11).

Although the changes in China’s posture are significant—and may enable shifts in strategy at a later stage—it is still premature to conclude that China’s nuclear strategy is changing. Rather than pursuing counterforce or limited nuclear options for coercive first-use, China appears to be building a more robust nuclear posture to firmly establish its second-strike capability in the face of what Chinese analysts perceive as a more threatening nuclear security environment. Although other factors—such as status concerns and a perceived need for a nuclear shield—may have influenced the expansion, the desire for a more robust second-strike force appears to be the main driver. Moreover, even if this posture represents a shift from the past and no longer aligns with the minimum deterrence position, it matches well with a pursuit of MAD. There is little evidence to suggest that China is seeking a posture that aligns with the delicate nuclear balance position.

Official statements indicate that Chinese leaders believe they require a more robust nuclear posture but provide no indications of a fundamental shift in nuclear strategy. In its most recent defense white paper from 2019, China maintained the long-standing formulation that it would maintain a nuclear arsenal “at the minimum level required for national security,” and upheld the no first use pledge (PRC State Council, 2019). In his report to the 20th Party Congress in October 2022, Xi Jinping stressed that China should “establish a strong system for strategic deterrence” but did not point to any shifts in strategy (Xinhua, 2022). In a recent statement, outgoing Central Military Commission Vice Chairman Xu Qiliang (2022) similarly stressed the importance of building a “high level” strategic deterrence system, but simultaneously underscored that China should adhere to “asymmetric balancing.”

Chinese civilian and military experts have in recent years argued that China faces a worsened nuclear security environment but have refrained from calling for a change in strategy. Prominent experts have argued that Chinese nuclear forces lack “redundancy,” which could incentivize US “nuclear opportunism” and threats (Li & Hu, 2018, p. 40). These and other experts therefore claim China needs to strengthen its deterrent forces, even if they have stressed the importance of quality over quantity (see also Luo, 2017). Some in China have called for a quantitative expansion, such as former editor Hu Xijin of the tabloid Global Times, who argued in 2020 that China should develop “1000 warheads, including 100 DF-41s” (Hu, 2020) to address the threat from the United States. Although Hu received pushback from Chinese experts, the fact that he was not censored indicates that arsenal size was under discussion already in 2018.

US missile defense efforts are the most prominent and long-standing concern of the Chinese experts, who fear that the United States could rely on its missile defense system to mop up any surviving missiles after a first strike (see, for example, Fan, 2018). In addition, Chinese observers have also long been concerned about US nuclear counterforce weapons, as well conventional precision strike capabilities (see Hu, 2021). These concerns have increased with the collapse of the Intermediate-Range Nuclear Forces Treaty, and the prospect of US deployment of land-based missiles in Asia, which both Chinese officials and analysts fear may impact the nuclear balance and pose a “severe threat” to “the survivability of Chinese nuclear forces” (PRC Ministry of Foreign Affairs, 2020; see also Chen & Ge, 2022; Guo, 2019). The US DOD (2022b) recognizes China’s “concerns about maintaining mutual nuclear vulnerability” (p. 159), and how they spring from fears of US advances in missile defense technology, hypersonic capabilities, long-range precision-strike missiles armed with low-yield nuclear warheads, conventional prompt strike, space surveillance assets and cyberweapons.

Beyond counterforce threats, Chinese analysts are also concerned about increasing US emphasis of lower-yield nuclear weapons, including through the deployment of the W76-2 nuclear warhead. Chinese analysts see the development of such weapons as part of a US effort to lower the threshold for nuclear use which was first reflected in the 2018 Nuclear Posture Review (Fan, 2018, p. 18). They further believe that as the balance of power in East Asia is tilting in China’s favor, the United States will increasingly compensate for conventional inferiority by relying on limited nuclear options (Luo, 2017, p. 49). Also here, the US DOD (2023) has recognized China’s increasing worries, highlighting that: “By late 2018, PRC concerns began to emerge that the United States would use low-yield weapons against a Taiwan invasion fleet” (pp. 111–112).

China’s silo-building efforts and broader nuclear modernization can credibly be linked to its concerns about its second-strike capability (Kristensen et al., 2023, p. 114). The silos willbolster China’s second-strike capability: Because the United States would have to strike a much larger number of targets in a first strike, its forces would be spread thinner, leaving fewer missiles available to target to Chinese mobile intercontinental ballistic missiles (ICBMs). Moreover, the new silo fields are located out of range for US conventional missiles. Prior to the building of the silos, Chinese experts associated mainly with the PLA Rocket Force Engineering University were debating how to make their land-based missiles more survivable. Although some stressed the benefits of mobility over silo-basing (Deng et al., 2018), others pointed to the importance of having a diverse force consisting of road-mobile, rail-mobile, and silo-based missiles, and argued that the “development and construction [of the silo-based force] must be stepped up to ensure effective deterrence” (Zhang et al., 2019; see also Dai et al., 2018). Similarly, the 2020 edition of the Science of Military Strategy, an important text published by the National Defense University, stated that while mobile missiles should be the “main direction” of China’s modernization, a mix of mobile and silo-based missiles “can complement each other’s advantages and increase the flexibility of strategic options” (Xiao, 2020, p. 383). Beyond the silos, other efforts that have caused concern such as the 2021 apparent fractional orbital bombardment system test can also be linked to Chinese worries about survivability. According to the US DOD (2022b), China is developing advanced hypersonic glide vehicles and the fractional orbital bombardment system “in part due to long-term concerns about US missile defense capabilities, as well as to attain qualitative parity with future worldwide missile capabilities” (p. 98).

China’s effort to bolster the readiness of its forces is also likely driven by survivability concerns. Whether China will adopt a launch-on-warning posture is still not a foregone conclusion: Although some military officers have called for such a posture, Chinese officials have also stated publicly that such a posture would violate its no-first-use pledge, and there is no clear evidence yet that a decision has been made (Hiim et al., 2023, pp. 170–171). China’s development of an early warning system, which is often seen as an indication that it is moving towards a launch-on-warning posture, could be used for other purposes, such as missile defense or bolstering the survivability of its nuclear forces. Moreover, if China were to shift to a launch-on-warning posture, it would represent a significant but not necessarily fundamental break with its past strategy: The main purpose of such a shift would likely be to strengthen survivability of particularly the silo-based forces and make it extremely challenging for the United States to maintain a damage-limitation capability (see Glaser & Fetter, 2017, p. 207). That said, if China adopted a launch-on-warning posture, it could have significant negative repercussions, harming crisis stability and leading to increased risk of accidental nuclear use (Eveleth, 2023, p. 53).

To be clear, there are parts of China’s nuclear modernization which are somewhat puzzling from a survivability perspective, and that may be driven by other factors. Although the JL-3 SLBM can reportedly reach the continental United States from within the first island chain—and thus allow China's ballistic missile submarines to stay in more protected waters—China’s JIN-class SSBN is allegedly noisy and therefore vulnerable. For this reason, some PRC scholars have even called for using the JIN-class as a training platform, and refrain from deploying them on deterrence patrols (Wu, 2017, p. 59). Nevertheless, China is continuing to construct additional JIN-class SSBNs beyond the six it has already built, while also developing a new class of SSBNs (US DOD, 2023, p. 108). Similarly, although the PLA Air Force may play a role in regional deterrence, Chinese bombers will not do much to bolster survivability vis-à-vis the United States. Finally, it is puzzling that China is fielding liquid-fueled, silo-based DF-5 ICBMs, which are likely less survivable than its newer solid-propellent missiles (see US DOD, 2023, p. 107).

A possible explanation for the development of systems with questionable survivability is the pursuit of great power status. In a recent study, Logan and Saunders (2023) point to bolstering its second-strike capability as a key likely driver of China’s modernization, but also find evidence that China’s development of a larger and more diverse arsenal could be driven by “prestige and status reasons” (p. 1). Another explanation they find support for is the “nuclear shield model,” where China seeks greater freedom of action for conventional military operations through a more robust deterrent, as well as to deter limited nuclear use by an adversary (Logan & Saunders, 2023; see also Denmark & Talmadge, 2021). Nevertheless, while the shield model is plausible, Chinese sources contain little direct discussion of such a goal (Fravel et al., 2023). Furthermore, as Fiona Cunningham (2023) has highlighted, “there is little evidence to support the possibility of tighter integration of China’s nuclear and conventional strategy, planning, or force building” (p. 12), which you would expect with a shield model.

In contrast, there is still little evidence to suggest that China’s nuclear expansion is driven by an effort to develop counterforce options and a first-strike capability. To be sure, nuclear weapons kept at high alert levels—as well as the fractional orbital bombardment system—could enable China to launch strikes against US nuclear targets with little advance warning. However, a true first-strike posture would require a massive Chinese expansion of its nuclear arsenal, with China obtaining at least parity—and preferably a significant numerical advantage—vis-à-vis the United States. As Jeffrey Knopf (2022) highlights, with China’s relatively smaller arsenal, it “is still nowhere near even a hypothetical first-strike capability against the United States” (p. 192; see also Logan & Saunders, 2023; Montgomery & Yoshihara, 2022, p. 50). Even if China were to develop 1000 nuclear weapons by the end of the decade, its arsenal would remain significantly smaller than that of the US. As mentioned above, there are no statements from Chinese leaders indicating that China is seeking to obtain a very large arsenal—rather, they are stressing the virtues of asymmetric balancing. If China massively expanded its fissile material, missile, and warhead production facilities, it would be an indication of China making a bid for nuclear superiority, (Logan & Saunders, 2023; p. 46: see also Talmadge & Rovner, 2023, p. 20). The US DOD has claimed China will likely rely on new fast breeder reactors to produce plutonium for its nuclear expansion, but other analysts have expressed skepticism, and highlighted how operating fast breeder reactors can be challenging (US DOD, 2023, pp. 109–110; Kristensen et al., 2023, p. 109; Podvig, 2023).

There is also limited evidence yet to suggest that China is endorsing a strategy of limited first nuclear use or coercive nuclear escalation. Despite concerns about the DF-26—which provides China with a nuclear precision-strike capability—a clear majority of the DF-26 missiles appear to serve in conventional roles (Kristensen et al., 2023, p. 109). A recent study indicated that China may have developed the DF-26 as a dual-use missile for cost-saving reasons (Logan, 2020, p. 36), and nuclear-armed versions of the missile may be intended primarily for regional deterrence against India and Russia (Cunningham & Fravel, 2019, p. 92). Moreover, despite the assertions by the US DOD, Kristensen et al. (2023, p. 124) highlight that there is no concrete evidence that China has or plans to field a low-yield warhead for the DF-26 or other missiles. Since China still lacks a large set of theater or tactical nuclear weapons (as well as advanced intelligence, surveillance, and reconnaissance capabilities), its posture is still not optimized for theater-level nuclear coercion (Cunningham, 2023, p. 10; see also Logan & Saunders, 2023).

Certainly, China’s attitude towards limited nuclear options could shift in the future, and there is some debate among Chinese experts about developing new weapons as a response to the increasing US emphasis on low-yield options. Although many experts in China have empathically rejected that it should develop low-yield warheads (e.g., Cui, 2018), others, such as retired PLA General Pan Zhenqiang (2018) argue that shifts in US policy have “enormous” consequences, and that China may need “more options for responding to a nuclear attack” (p. 129), including non-strategic weapons. Along similar lines, Tong Zhao (2021) has speculated that China could seek to rely on the DF-26 for proportional nuclear retaliation at the regional level if the United States resorts to limited first use. In the event of further escalation, China could rely on intercontinental-range, highly precise missiles for limited strikes. In other words, although China does not yet appear to have developed dedicated limited nuclear options yet, increasing fear of US employment of such weapons makes a shift more likely. Although China might initially develop dedicated limited nuclear options as a response option, it would represent a shift from its current strategy, and other states in the region will perceive them as tools of coercion.

In sum, even if China is expanding and modernizing its arsenal, these steps likely reflect a shift towards deterrence based more firmly on a secure second-strike capability, or even MAD. China’s nuclear forces are still not nearly as large, robust and sophisticated as the delicate nuclear balance scholars would advocate, nor is there much evidence to suggest that China is moving towards building such an arsenal. The most likely future shift in China’s approach to nuclear weapons would be China adopting a launch-on-warning posture. It is also possible that China could develop and deploy a low-yield warhead for proportional retaliation, even if there is limited evidence of this yet. Although both shifts would be concerning, they would not necessarily constitute a break from the prescriptions of the mainstream nuclear revolutionaries.

Conclusion

Recently, scholars have criticized the nuclear revolution theory and highlighted how leaders in the US and the Soviet Union continued to see the balance of terror as delicate. However, the revolutionary spirit appears to endure in Beijing. As this article demonstrates, Chinese leaders’ past beliefs that a small and potentially vulnerable arsenal is sufficient for deterrence closely matched the position of a minority of nuclear revolution scholars, such as Kenneth Waltz. With its recent nuclear expansion, China appears to be turning into more of a mainstream revolutionary, basing its deterrence policy on a secure second-strike capability or even assured destruction forces. So far, the expansion of its arsenal appears consistent with an effort to bolster the arsenal’s survivability in the face of what many in China see as increasing threats from US missile defenses and conventional and nuclear counterforce capabilities. Concerns over status and prestige, and the desire for a stronger nuclear “shield” are other possible drivers, but there is limited evidence that China is seeking an arsenal suited for coercive nuclear escalation, and even less evidence for a pursuit of counterforce capabilities and a first-strike option.

To be sure, current shifts in China’s nuclear posture could enable future changes in strategy, and overall, China’s nuclear trajectory is more uncertain. Most notably, China’s attitude to limited nuclear options might be shifting, and there appears to be some debate in the Chinese community of nuclear experts about whether to develop non-strategic nuclear weapons as a response option to counter the threat of US first-use of low-yield weapons. If China deploys such weapons, it would represent a break with its past strategy, and likely be perceived as a tool of coercion among its neighbors. Nevertheless, if deployed in small numbers, it would not necessarily break with mainstream renditions of the nuclear revolution theory, which does not altogether reject the utility of limited nuclear options. A move in the direction of a delicate nuclear balance position would entail much more substantial changes, most notably a large buildup of counterforce capabilities, and the deployment of a large and diverse set of limited nuclear options. China’s posture is not—despite the buildup—anywhere near a first-strike capability, nor is it optimized for coercive nuclear escalation campaigns. Rather, it is starting to approach an assured destruction capability.

Certainly, one may debate whether greater confidence among Chinese leaders in their second-strike capability is favorable and will reduce the chances of conflict in Asia. According to many nuclear revolution scholars, such confidence should limit the chance of war. If Chinese leaders believe strongly in the stability-instability paradox, greater nuclear confidence could increase the risk of conflict (see Christensen, 2012; Snyder, 1965; Talmadge & Rovner, 2023). Arguably, Asia’s maritime geography at the outset makes the risks of a limited war more acute than in the Cold War European theater, where the stakes were higher, and a conflict would probably have escalated rapidly (Tunsjø, 2018, pp. 141–144). Because of geopolitics, the stability-instability paradox could thus be more acute in the new superpower rivalry between the US and China. However, their increasing concern about US limited nuclear use in particularly a Taiwan conflict indicates that Chinese leaders are becoming somewhat less confident that a conventional conflict would not lead to US nuclear use (Hiim et al., 2023, p. 160; cf. Cunningham & Fravel, 2019).

China's nuclear policy is more than a single case in the debate between the nuclear revolution theorists and their critics. Whereas delicate nuclear balance scholars rightly argue that the Soviet–US nuclear dyad represented “the canonical case” of 20th century (Lieber & Press, 2020, p. 41), the China–US deterrence relationship will be the fulcrum of the 21st century. If China continues to adhere to a nuclear strategy that is more restrained than the strategies of the Cold War superpowers, future great power rivalry will—at least in this crucial respect—play out very differently from that of the not-so-distant past.

At the same time, the findings of this article indicate that US policy choices will in no small part influence the future of China’s nuclear posture. There is currently a major debate in the US about how to meet China’s nuclear expansion, with many calling for a shift in US posture. A study group convened by the Center for Global Security Research at Lawrence Livermore National Laboratory (2023) recently argued that in the face of a “two-peer problem,” where the United States has to deter Russia and China simultaneously, the United States should “plan and prepare to deploy additional warheads and bombs” (p. 7). Similarly, the bipartisan Congressional Commission on the Strategic Posture of the United States (Creedon et al., 2023, pp. 35, 48) recently recommended several changes to US nuclear forces, including preparing to upload warheads in reserve and deploying additional theater-range weapons with variable yield options—non-strategic weapons—to the Indo-Pacific region. Unfortunately, such steps will likely generate a response from China, such as a further increase in its nuclear stockpile, or even a decision to field new and potentially destabilizing weapons, such as dedicated limited nuclear options or fractional orbital bombardment systems. This, in turn, would further fuel arms race dynamics.

This article’s findings also hold implications for international relations scholarship more broadly. It illustrates the potential for bias by generalizing based on US experiences (Braut-Hegghammer, 2019). Critics of the nuclear-revolution theory explore other cases (most notably that of the Soviet Union), but the gist of their empirical work centers on the nuclear policies of the United States. This creates a risk of scholars projecting their own assumptions, based on US experiences, onto China’s nuclear policies and that of other nuclear powers—a tendency other scholars have observed in discussions about China’s broader military strategy (Fravel & Twomey, 2014). Arguably, the concerns about China threatening limited first-use or even all-out first-strikes could be a projection of such assumptions: First use is still at odds with China’s declaratory policy and appear inconsistent with Chinese strategists’ beliefs about nuclear escalation and deterrence more broadly. By making such projections, analysts and policymakers may contribute to feeding arms race dynamics.

Acknowledgements

For helpful comments and suggestions, the author thanks M. Taylor Fravel, Robert S. Ross, Øystein Tunsjø, and the anonymous reviewers.

Disclosure statement

No potential conflict of interest has been reported by the authors.

Additional information

Funding

Research for this article was supported by the Security in Asia program at the Norwegian Institute for Defence Studies.