A Multipolar Nuclear Asia in the Trump Era

On December 22, 2016, President-elect Donald Trump used his electronic bully pulpit to opine on nuclear weapons policy. “The United States must greatly strengthen and expand its nuclear capability until such time as the world comes to its senses,” he tweeted, seemingly unprovoked by any global happening.¹ He likely was reacting to a series of meetings with US Air Force and US Navy uniformed leadership the previous day, though thematically he was restating beliefs he had offered in the 2016 presidential campaign and may have held for several decades.² In the face of media commentary questioning the wisdom of such a strategy, Trump telephoned a popular morning television show to underscore his commitment to such a path even if it generated countervailing moves by US adversaries. If US foes wanted to compete with a US nuclear expansion, he stated “Let it be an arms race. We will outmatch them at every pass and outlast them all.”³

Trump’s extemporaneous nuclear remarks may turn out to be bluster—especially as he attempts to fulfill prior promises to cut taxes and expand conventional military forces. At a minimum, it seems that Trump’s nuclear predilections strongly differ from President Obama’s, who rhetorically reemphasized the US commitment to global nuclear disarmament in his first months in office in Prague and his last year in office in Hiroshima.

This stark shift in nuclear preferences within US leadership occurs alongside a longer structural evolution in the international system from bipolarity to unipolarity and now to multipolarity. Nowhere is that shift more visible than Asia where the rise of China has been followed by the (somewhat slower) rise of India and may spur the return of a more normal Japan. This multiplicity of potential reacting states has led many to speculate US nuclear steps may trigger, in Matthew Kroenig’s analysis, “a chain reaction that encourages several states to increase the size and sophistication of their nuclear arsenals.”⁴ Such fears are not new. Michael Krepon worried at least as early as 2003 that strategic posture changes by the George W. Bush administration could cause a “nuclear cascade” that might prompt “further growth in nuclear stockpiles” and “exacerbate spikes in regional tensions.”⁵

If Trump chooses to “greatly strengthen and expand” US nuclear forces, how likely is it that multipolarity will amplify the effect of that decision, triggering a multisided arms race that is resistant to old treatments? This article offers a note of modest optimism based on the special roles played by China and India in any potential Asian nuclear cascade. The specific geography of the Sino-Indian competition may combine with ideational and organizational inertia to attenuate the effects of any Trump-induced nuclear buildup. This analysis does not suggest such an initiative is prudent or beneficial for US interests, but it does suggest the most probable consequences of such a course are merely injurious and not unavoidably catastrophic.

The Theoretical Instability of Multipolarity

The concerns with stability in multipolar orders are intuitive. States must face multiple potential competitors, and they may have to face combinations of those competitors. If states tend to gravitate toward parity in nuclear stockpiles, as a rule, then multipolarity would make it impossible to maintain a stable equilibrium, since a state would need to be able to possess sufficient nuclear weapons to equal the force of multiple states in combination, meaning a state would need more nuclear weapons than any one potential adversary state. As each state increases the number of weapons to deal with potential combinations of foes, those same adversaries would be compelled to react with offsetting increases of their own. Participants are trapped in an uncontrolled arms race. Additionally, such an arms race is essentially uncontrollable if parity is the goal. Even if expansive multilateral treaties that codified unequal arsenals were possible, such as the Washington and London naval treaties of the 1920s and 1930s, the possibility of changing combinations of foes would still make stable equilibrium unachievable.⁶

Compare that situation with that of a bipolar competition. States can tacitly and explicitly bargain toward parity, and can essentially settle on a negotiated settlement at any level they assess is consistent with a survivable second-strike force. Without multiple independently targetable reentry vehicle (MIRV) capability, such parity is fairly stable. Any potential disarming (counterforce) strike launched by the initiating state also has the advantage of vastly reducing the number of potential retaliatory targets for the unfortunate recipient state. At rough parity, a counterforce attempt requires the initiating state to use nearly the entirety of its force, while the attacked state can focus its retaliation on counter-value targets in response. The situation is more complicated with the introduction of MIRVs, but bipolar competitions remain vastly simpler in theory than their multipolar counterparts.⁷

The Observed Stability of Nuclear Multipolarity in Asia

The multipolarity-induced instability described above is derivative of beliefs that parity is required for nuclear deterrence. In Asia, however, two nuclear states have concluded that such a quest for parity—let alone nuclear advantage—is unnecessary and wasteful. Instead, India and China have both embraced assured retaliation strategies that emphasize maintaining survivable second-strike forces with little emphasis on the relative size of the nuclear arsenal in comparison with nuclear foes.⁸ Importantly, Indian and Chinese decision-makers concluded that so long as there was some reasonable possibility that a portion of their nuclear arsenal might survive a nuclear first strike, no other nuclear weapons power would ever conclude that the political benefits of such an attempt were worth the risk. Nuclear deterrence, in their estimation was and is fairly easy to achieve against a wide-range of potential adversaries. This is not to say that adversary force structure and planning played no role in their own sizing and doctrinal decisions, but rather that the nuclear balance was comparatively “unimportant” in their estimation and that they did not view nuclear deterrence as “delicate.”⁹

The evidence of this relatively relaxed view of deterrence was starkest during the Cold War when China accepted vast inferiority in its nuclear arsenal compared to both of its potential nuclear competitors (the Soviet Union and the United States) and India contented itself with a recessed deterrent following its 1974 “Peaceful Nuclear Explosion.” China did not have an operational ballistic missile capable of targeting the capitals (Washington and Moscow) of its two principal adversaries until 1981 (seventeen years following its first nuclear weapons test), while India did not have an operational air delivery capability until approximately 1995 (twenty-one years following its first nuclear weapons test)—and it did not flight test an intermediate-range ballistic missile capable of hitting eastern Chinese cities until 2006.¹⁰

This earlier lethargy could have been ascribed to resource limitations. But even in the recent period, when China and India have recorded spectacular economic growth, which has enabled substantial increases in both nations’ defense budgets, there has been no rush toward parity. As Figure 1 (above) shows, China today still likely deploys only one nuclear weapon for every eighteen the United States holds in its stockpile. While this ratio is substantially better than the nearly 36:1 ratio that China faced in 2005, nearly all that improvement has come from US force reductions rather than Chinese force increases. Similarly, while India now likely possesses a nuclear arsenal only half the size of China, compared to an arsenal one-sixth as large in 2005, it has seemingly accepted Pakistan having a modest quantitative edge in its nuclear stockpile despite Pakistan having one-seventh the population of India and an economy one-eighth the size of India’s. Even without examining the details of delivery vehicles, this behavior is not consistent with that of states seeking parity and certainly not that of states jockeying for nuclear advantage.

This continued nuclear nonchalance is even more evident once delivery vehicles are considered. Of China’s approximately 260 warheads, less than one hundred sit atop missiles capable of reaching the United States, even when generously counting missile launchers that can reach Alaska but not the continental United States.¹¹ Similarly, India even today appears to have only a handful of operational Agni-3 missile launchers, capable of ranging at least some major Chinese cities, and may have no operational Agni-4 or Agni-5 missile launchers capable of targeting Beijing and Shanghai, even though flight-testing of those longer-range systems suggests they are nearly ready for deployment.¹²

Why?

The explanations for this observed stability can be grouped into three clusters: organizations, ideas, and threats. These explanations are not mutually exclusive and have operated in a reinforcing fashion to produce the historical pattern thus far.

Ideas: Unlike conventional wars, nuclear wars have mercifully not been fought. There is no observed experience from which observers can learn the political effects of two-sided nuclear conflict. The young RAND analyst Alain Enthoven chastised a smug US Air Force officer by correctly observing, “General, I have fought as many nuclear wars as you have.”¹³ At most there are hints of the political effects of nuclear weapons from past crisis behavior. Different policymakers operating in distinct contexts unsurprisingly may arrive at quite divergent beliefs about the nature of nuclear deterrence.¹⁴ There is no shared factual basis that will lead such policymakers to converge on the “correct” answer. Peter Lavoy refers to whatever beliefs policymakers adopt or attempt to promulgate as “myths” precisely because they are “unverifiable.”¹⁵

Ideas about nuclear weapons proved long-lasting in both India and China. Lavoy argues that foundational myths set in place by India’s first prime minister Jawaharlal Nehru and nuclear scientist Homi Bhabha continued to shape India’s nuclear program long after both of their deaths in the mid-1960s.¹⁶ Relatedly, Andrew Kennedy points to Nehru’s commitment to global disarmament as limiting India’s nuclear choices during Nehru’s lifetime.¹⁷ Similarly, M. Taylor Fravel and Evan Medeiros emphasize the belief, strongly held by both Mao Zedong and Deng Xiaoping, that nuclear weapons were “tools for deterring nuclear aggression and coercion, not… tools to be used in combat to accomplish discrete military aims.”¹⁸ Given Mao and Deng’s long leadership tenures, effectively ruling for the first forty years of the Communist Chinese regime, they left an enduring imprint on Chinese strategic thinking and doctrine.

Organizations: In the Chinese case, Fravel and Medeiros additionally point to an organizational structure that segregated and isolated China’s nuclear strategists, militating against serious rethinking of the ideas articulated by Mao and Deng until at least the late 1980s. Even when organizational strictures reduced, and Deng faded from the scene, Chinese debates about nuclear doctrine largely ended up reaffirming Mao and Deng’s core ideas. Fravel and Medeiros suggest that the opening of nuclear discourse to multiple elements in the military and defense scientific community may counterintuitively hamper the emergence of a new strategy that breaks with the past. Old ideas may persist as a multiplicity of veto players each push for their own nuclear vision but reject those of bureaucratic opponents.¹⁹ The resulting situation favors ideational inertia.

Itty Abraham argued in the Indian case that there, too, existed a strategic enclave insulated from outside pressures, especially pressures from the Indian military. This enclave maintained an interest in the nuclear weapons program in part to justify continued funding despite the failure of this community of Indian government scientists and engineers to achieve other potential missions, such as the provision of civil nuclear energy or the development of an indigenous conventional weapons industry.²⁰ Gaurav Kampani argues that the strategic enclave maintained considerable autonomy because of the desire to shield nuclear activities from international observers given the pressures of the nuclear nonproliferation regime. This slowed development of the nuclear weapons program, creating capabilities more consistent with a relaxed view of nuclear deterrence.²¹ Even after the overt 1998 nuclear tests reduced the need for secrecy and permitted subsequent political steps to increase the external oversight of the Indian nuclear weapons program, Indian civilians’ lingering fear of the military appears to limit the degree of debate permitted internally. An assured retaliation doctrine, which provides for a limited role for the military as nuclear custodians, appears more consistent with Indian civil-military relations than a more expansive doctrine that required larger nuclear forces with higher levels of operational readiness.²²

Threats: In both the Indian and Chinese cases, unique organizational heritages contribute to stasis or evolutionary change in nuclear doctrine. These bureaucratic and domestic political factors have greater heft because the threat environment surrounding both Asian giants does not necessitate nuclear first strikes for defense. Vipin Narang proposes three measures to determine the threat environment faced by a regional nuclear power. Does it face a conventionally superior adversary? Does it share an unimpeded land border with that adversary? Can that adversary launch an offensive attack that will not be defeated through conventional defense-in-depth? In conventionally inferior states, the threat environment overwhelms concerns about domestic civil-military fights and encourages nuclear strategies characterized by nuclear first use. In the absence of such threats, regional nuclear powers are mostly content with assured retaliation strategies, Narang argues.²³ This empirical regularity is important because asymmetric use strategies often require large arsenals for credibility—as even a cursory review of the US debate about extended deterrence makes clear.

The Unpersuasive Justifications for Chinese and Indian Arms Buildups

Nuclear beliefs, insulated organizational arrangements, and manageable threat environments have led to comparatively small Chinese and Indian nuclear arsenals. If Beijing or New Delhi were to revisit existing force postures, they would need some justification for additional weapons systems. When potential justifications are examined, however, they fail to withstand closer inspection.

Nuclear weapons carry limited ability to facilitate aggression in states that do not face a conventionally superior neighbor.²⁴ To some extent there may be a “stability-instability paradox,” that is stable nuclear deterrence might permit conventional tomfoolery. For conventionally weaker states, limited war becomes “cheaper” because stronger states cannot respond to limited conventional aggression with full-scale war because full-scale war increases the possibility of nuclear escalation. For states like India or China, however, that already are already able to mount a convincing defense against conventional territorial aggression, there is little added benefit of threatening nuclear first use to stop conventional intrusions. Their armies can already stop major territorial grabs, and nuclear use is incredible in the face of limited territorial aggression. This condition does not always obtain. Russia may have something to fear from a Chinese conventional fait accompli. It might not be able to take the territory back with conventional force alone. It might be attracted to nuclear weapons to deter conventional defeat for the same reasons that Pakistan is in the South Asian context. Even here, the theoretical danger to Russia is minimized by the practical reality that China would likely only seize Russian territory to gain energy resources that thus far Russia has proven quite content to sell to its neighbor.

Similarly, attempting to pursue damage limitation as a strategy seems foolish in the specific contexts in which China and India find themselves. Damage limitation strategies utilize attacks against adversary nuclear forces and associated command and control capabilities (perhaps alongside missile defenses). They seek to degrade or defeat adversary nuclear use, rather than deter it.²⁵ For New Delhi or Beijing to transition to damage limitation would almost certainly require both countries to abandon their commitment to no first use of nuclear weapons. At a minimum it would require transitioning to a posture of launch on warning. Given relaxed states of readiness during peacetime for much, if not all, of the Chinese and Indian nuclear forces, this would require substantial changes in either nuclear force. ²⁶ For states, like Pakistan today or the United States during the Cold War, that envision nuclear first use on the battlefield, damage limitation strategies are appealing complements. For states that have settled for assured retaliation, damage limitation strategies are a big intellectual leap.

There are individuals in both countries who advocate for the abandonment of no first use (especially in India) or higher states of peacetime readiness (in China).²⁷ They have failed in persuading their peers and superiors to pursue more ambitious nuclear strategies in part because of the domestic politic reasons outlined above, but also because the incentives continue to dissuade damage limitation strategies. As a first order consideration, Indian civilian leaders have shut down Indian military personnel advocating for a first use strategy by arguing that it is always better to have some chance of avoiding a nuclear exchange than guaranteeing through first use that such an exchange take place.

Additionally, the specifics of the Asian nuclear competition are also different from those that obtained in the Cold War superpower cases, even setting aside the issue of multipolarity. The comparatively shorter distances involved have permitted the widespread use of mobility as a survivability solution for land-based missiles. While land-based mobile missiles are somewhat more vulnerable than their submarine-launched equivalents, they do confer quite a bit more survivability compared to silo-based weapons. In India, China, and Pakistan these land-based mobile missiles are being supplemented by established air-delivery options in the India-Pakistan dyad and nascent submarine-launched options in the India-Pakistan and India-China dyads. While neither India, Pakistan, nor China engage in airborne alert for nuclear forces or maintain constant submarine deterrent patrols, they would have these options in a crisis. They would also be able to disperse land-based missiles. The real-time intelligence, surveillance, and reconnaissance task of identifying all of the peacetime storage locations for land-based missiles as well as all potential launch sites likely exceeds by a good margin current US abilities, and vastly exceeds the capabilities of all of the regional states. A leader would have to be incredibly risk-acceptant and willing to launch a full-scale nuclear strike “out of the blue” in order to have any confidence in a damage limitation attack. And the attractiveness of this option would be reduced further by the knowledge that, even if successful, partners of the attacked state would still be standing with full nuclear arsenals. In other words, China would continue to exist unscathed even after an Indian nuclear attack on Pakistan, while the United States would still be around even after a Chinese nuclear salvo against India. Damage limitation attempts would place even a successful first-launcher in a position of extreme vulnerability in their other international competitions. That is even setting aside the near certainty that the attacked state would retain at least a modest retaliatory capability, which would inflict tremendous damage on the attacker, even if that damage might merely be crippling rather than society-ending.

Asymmetric force sizes and asymmetric distances create another stabilizing force in any potential multipolar nuclear arms race. Pakistan is expanding fissile material at perhaps the fastest rate of any country on the planet. However, an increasing portion of the Pakistani force appears to be earmarked for battlefield missions.²⁸ India does not need to match Pakistan 1:1 in responding to this buildup, since if Pakistan were to use nuclear weapons on the battlefield, there is little reason for India to respond proportionally on the battlefield. Even if India wants to avoid escalating to the use of nuclear weapons against cities, there are ample military targets that India could attack in response to Pakistani “limited” nuclear use. To offer a specific example, if Pakistan were to use twenty nuclear devices to stop an Indian armored advance, India could counterattack by using two nuclear weapons against bases without losing any credibility or coercive leverage in the process. This dramatically dampens India’s need to respond to any given increment of a Pakistani nuclear buildup. Even if India came to believe it needed to engage in nuclear warfighting on the battlefield to deter Pakistani first use, it would invest in shorter-range systems that would do little to endanger Chinese population centers (excluding perhaps Lhasa).

What India currently lacks is missiles capable of holding China’s eastern seaboard at risk, while India possesses extant but limited capabilities to threaten population centers in central China. Today India may have no deployed systems capable of threatening Beijing or Shanghai, and only manned aircraft and Agni-2 and -3 missiles capable of threatening moderately sized cities such as Chengdu, Chongqing, and Kunming or major infrastructure targets such as the Three Gorges Dam. Currently without MIRVs, this means, perhaps, as few as 12 missile-delivered warheads and 48 aircraft-delivered weapons.²⁹ The ability of Indian aircraft to survive extremely long distances over Chinese territory is uncertain at best. In other words, any Indian modernization effort in the near- to medium-term will likely attempt to hold at risk China’s eastern seaboard. Such modernization does not increase the threat to Pakistan in any meaningful way since India already has several dozen additional missiles capable of reaching all major Pakistani population centers. Pakistan only has ten cities of over 1,000,000 people and fewer than 30 cities with more than 300,000 people.³⁰ Pakistan’s deterrence calculations are not that different if India has 100 weapons, as it likely does today, or 200 or 300 weapons.

Similarly, what China lacks is long-range ballistic missiles capable of reaching the continental United States. China has hundreds of systems—possibly including hundreds of cruise missiles—capable of reaching most Indian cities. New Delhi, after all, is less than 450 kilometers from the Sino-Indian border. If China increases the number of long-range systems capable of reaching the United States in its arsenal, it does not matter much to India.

Pushing on either end of this Asian “chain” does not lead to ripples along its length. If Pakistan increases its nuclear arsenal to several hundred weapons, India only needs to respond modestly (if at all) and doing so does not meaningfully threaten China. If China needs to increase its long-range deliverable systems by 100 or 200 systems, India’s strategic circumstance does not qualitatively change.

A US Arms Buildup Is Still Unwise

The above analysis suggests cautious optimism about the consequences of a US arms buildup for a nuclearized multipolar Asia. This does not make such a policy prudent or risk free. First, it is costly. Jon Wolfsthal and colleagues estimate that over a thirty-year window modernizing the triad of US nuclear forces could cost USD 850 billion to USD 1.1 trillion, or USD 28 billion to USD 36 billion per year. The bulk of these costs would be incurred from 2024 to 2030, rather than being distributed across the three-decade timeframe. This budgetary “modernization mountain” is the cost of merely maintaining and updating the force, not engaging in a buildup or arms race.³¹ Even if the only goal of US defense spending was to deter China, it is difficult to imagine that spending even more money on the nuclear deterrent would be the best marginal use of resources.

Second, one important contributor to the modest Indian and Chinese programs has been relaxed beliefs about the requirements of nuclear deterrence. These beliefs persisted in part because of organizational isolation and secrecy, but for different reasons that bureaucratic insulation has diminished for both the Indian and Chinese strategic enclaves. If the United States argues loudly and consistently for many years that its already large arsenal is insufficient for deterrence tasks, especially if it appears that it seeks to engage in damage limitation against China, it may lead to a reconsideration in Beijing and New Delhi about the requirements of deterrence. The vast disparity between US and Chinese nuclear programs make a Chinese “sprint to parity” unlikely, but if Washington continues to argue that its overwhelming nuclear advantage over China is inadequate it might encourage more vigorous Chinese competition. Washington would be saying that present expenditures are not enough even for China’s small arsenal, but additional expenditures will be sufficient. Even if Chinese force planners believe Washington is foolhardy, they might wish to ante up additional resources to deprive their American counterparts of greater confidence.

Third, and relatedly, this perceptual competition can lead to unnecessary expenditure even if the structural incentives do not favor arms racing. During the Cold War, some US policymakers believed that under mutually assured destruction, the details of the US-Soviet nuclear balance did not matter directly, but that the nuclear balance did affect international political outcomes through the perceptions it generated among international audiences.³² Chinese strategic thinkers may not buy US arguments about the importance of the nuclear balance for crisis behavior, but they could worry that Japan or South Korea or India might come to believe such arguments. If US allies and partners became more risk-acceptant due to an unfavorable nuclear balance, China might have incentives to compete even if it thought the competition was irrational.

It remains to be seen whether to take President Trump’s foreign policy pronouncements either literally or seriously. They may prove to be bluster. There is modest comfort, however, in the fact that the circumstances in Asia are likely to dampen, rather than worsen, any US nuclear weapons initiatives in the near- to medium-term. Multipolarity is not destiny. It is mediated by the ideas, organizations, and threats faced by its constituent members, as well as the geography on which they are arrayed.

* Thanks to Keith Preble for his research assistance