Table of Contents

Any attempt to mitigate the risks of inadvertent escalation generated by the growing entanglement of non-nuclear weapons with nuclear forces and their enabling capabilities must begin with a serious effort to understand these risks. Given the extent to which they depend on perceptual factors—what one side in a conflict perceives its adversary’s intent to be—it is critical that Washington understands Moscow’s and Beijing’s concerns, and that Moscow and Beijing understand Washington’s (whether or not those concerns are viewed as fair or reasonable).

In this context, perhaps the single most important observation from the preceding chapters is that, within traditional and contemporary strategic thought in both China and Russia, very little attention has been given to the possibility that escalation might be unintended. It is quite unusual for experts from these countries to express serious concern about inadvertent escalation, as the authors in this volume do (even if Tong Zhao and Li Bin stress that they are somewhat more sanguine than many of their Western counterparts). The prevailing assumption in both China and Russia is that any increase in the level of violence in a conflict would be deliberate. Alexey Arbatov, Vladimir Dvorkin, and Petr Topychkanov, for example, describe “a visceral assumption among contemporary Russian strategists that the decision to use force—including nuclear weapons—would be a rational step.” In a similar vein, Zhao and Li note the existence of a

traditional Chinese belief that the course of a war can be well controlled and managed by top commanders. The various uncertainties associated with waging a war or the possibility that top commanders may not fully understand the situation or be able to effectively control military operations has not been seriously considered.

Ironically, the belief that inadvertent escalation is improbable actually makes it more likely because political and military leaders are left less inclined, in peacetime, to take steps that could mitigate the risks, and more inclined, in wartime, to interpret ambiguous events in the worst possible light.

Moscow and Beijing share the assumption that the United States seeks to undermine their nuclear deterrents with advanced conventional capabilities. As a result, U.S. non-nuclear operations that inadvertently implicated either of their nuclear forces would risk being interpreted as the beginning of a conventional counterforce campaign. In a major conventional conflict, such operations would be worryingly likely. Arbatov, Dvorkin, and Topychkanov, for example, note that “Russian strategic submarines and bombers are kept at the same bases as general-purpose naval vessels and aircraft, and [non-nuclear] strikes designed to target the latter might inadvertently destroy the former.”

Even more acute risks of escalation might stem from non-nuclear attacks against dual-use enabling capabilities, especially early-warning satellites. In this regard, both sets of authors are unusual in recognizing that it is not just U.S. attacks against China or Russia that could prove escalatory—Chinese or Russian attacks against the United States could do so too.

Zhao and Li, for example, note calls within the Chinese strategic community to attack U.S. early-warning satellites in a conflict in order “to degrade U.S. theater missile defense capabilities and hence to ensure the efficacy of Chinese conventional missile strikes against American and/or Taiwanese targets in the region.” They go on to argue that

although some Chinese experts understand that these satellites also provide strategic early warning of an incoming nuclear strike, these experts seem to expect the United States to be able to correctly interpret the use of [anti-satellite] weapons in a war that is conventional, limited, and regional. They reason that because China appears to have no capability to undermine the United States’ massive nuclear forces, it would make no military sense for China to even try to do so. However, they neglect the possibility that the United States might interpret such strikes as preparations for the first use of nuclear weapons designed to scare rather than disarm.

That said, the focus in both analyses is escalation risks generated by U.S. capabilities. In this context, both sets of authors see trouble brewing in the form of various technological developments, in hypersonic boost-glide weapons in particular. According to Zhao and Li, Chinese experts “have always assumed that the United States is interested in deliberately using hypersonic weapons to preemptively attack China’s nuclear forces,” and that, as a result, Beijing would be “more prone to interpret an ambiguous event as an attack against its nuclear forces.” (Unlike ballistic missiles, boost-glide weapons are maneuverable so their targets are not clear until impact. This characteristic could generate ambiguity in the event the United States launched such a weapon at a region of China containing both nuclear and non-nuclear forces.)

Arbatov, Dvorkin, and Topychkanov, meanwhile, focus on the challenges of detecting an incoming boost-glide attack. Early detection is particularly important for Russia, which still leans heavily on launch-under-attack as a way of enhancing the survivability of its intercontinental ballistic missile (ICBM) force. Arbatov, Dvorkin, and Topychkanov argue that

ground-based radars would only detect an incoming glider late in flight—too late, in fact, to launch ICBMs before they were hit. As a result, a launch-under-attack option would have to be executed exclusively on the basis of satellites’ detecting the launch of boost-glide weapons, without confirmation of an attack from ground-based radars.

The undesirability of Russia’s adopting a policy of launching ICBMs on the basis of a single detection technology does not need to be spelled out. Over time, China may face pressure to adopt a similar policy. Zhao and Li note that there are calls within China to adopt a launch-on-warning strategy, and that China is developing the technology that would enable it to do so. If China does make this shift, it will face the same challenges as Russia in preparing plans to counter a U.S. boost-glide attack.

That said, Russia and China have very different nuclear force postures today, and there are also important differences in the character of the conventional war that each might wage against the United States. As a result, the escalation dynamics in a U.S.-Russia conflict could have important differences from those in a U.S.-China conflict. Understanding the peculiarities of each scenario is critically important.

At the center of contemporary Russian strategic thought is the concept of an “air-space war” against the United States and the North Atlantic Treaty Organization. While the meaning of this concept may, as Arbatov, Dvorkin, and Topychkanov argue, be frustratingly elusive, an air-space war is viewed “as a prolonged endeavor involving an integrated technological and operational continuum of nuclear and non-nuclear operations, defensive and offensive capabilities, and ballistic and aerodynamic weapons.” Such a conflict, they argue, would create “a breeding ground for entanglement.”

Russia’s nuclear posture contributes to this entanglement. Russia has a large force of tactical nuclear weapons, which it might employ early in a conflict. Arbatov, Dvorkin, and Topychkanov assess that these weapons “might accidentally be attacked . . . because their delivery vehicles are collocated at bases with—and can be used together with—general purpose forces and weapons.” Even more worryingly, entanglement might lead Russia to conduct “limited strategic strikes” in an effort “to thwart U.S. naval and air forces that were engaged in a conventional conflict and perceived as conducting a conventional counterforce offensive by launching attacks against airfields, naval bases, and their C3I [command, control, communication, and information] facilities.” Limited strategic strikes, it should be emphasized, are not part of Russia’s official declaratory policy, but they are openly advocated by government-affiliated experts, and Western analysts (or at least those without access to classified information) do not appear to have picked up on these discussions.

China has a much smaller and less diversified nuclear force than Russia, and has pledged not to use nuclear weapons first. While alleviating some risks of entanglement, this posture may simultaneously exacerbate others. For example, because China’s nuclear force is much smaller than Russia’s, Beijing may be even more concerned about the possibility of conventional counterforce—even if it is vanishingly unlikely that China would use nuclear weapons to attack the U.S. non-nuclear forces that Beijing believed were threatening its nuclear deterrent. Importantly, Zhao and Li point out that because of mistrust between the United States and China, the potential benefits of China’s nuclear posture are less marked than they might otherwise be. In particular, Chinese experts generally have “complete faith” in their country’s no-first-use commitment. As a result, they tend to believe that Washington would not interpret ambiguous Chinese actions, such as attacks against early-warning satellites, as preparations for nuclear use. Recognizing that many Western experts “challenge the unconditionality of China’s no-first-use commitment,” Li and Zhao believe that there is significantly greater scope for misunderstanding.

The character of the escalation risks that might arise in a conflict is further shaped by the strategic geography of the theater of operations. Naval operations, which would be an important adjunct to a land war in Europe, would be at the very center of a U.S.-China conflict.1 In this context, dangerous interactions could arise between U.S. unmanned underwater vehicles (UUVs) and China’s nascent force of ballistic missile submarines (SSBNs). Zhao and Li note Chinese concerns that, in the near future, “UUVs could be deployed at the entrance to an enemy’s submarine base or near a maritime chokepoint to track and trail submarines.” Not only might such operations threaten both Chinese SSBNs and attack submarines, but “even if the United States wanted to threaten only China’s attack submarines and not its SSBNs, there would be a real risk that China would nonetheless suspect that its sea-based nuclear deterrent capabilities were in danger.” This form of entanglement may not be unique to UUVs; U.S. attack submarines could also threaten both Chinese SSBNs and attack submarines. However, UUVs compound the problem—not least because they could potentially be deployed in much larger numbers than attack submarines, and because, without a crew to worry about, UUVs could be tasked with more aggressive but riskier operations than attack submarines generally undertake.

One important—and difficult to explain—difference between Russian and Chinese perceptions concerns the plausibility of cyber attacks against nuclear and dual-use C3I systems. Zhao and Li observe that “Chinese analysts have demonstrated an acute awareness of the potential vulnerabilities of the country’s nuclear C3I system, particularly against cyber infiltrations.” They explore how, in light of this perceived vulnerability, the discovery that a nuclear C3I network had been penetrated could be highly escalatory—even if the attacker’s goal was no more malign than espionage. By contrast, Arbatov, Dvorkin, and Topychkanov are less concerned about the escalation implications of cyber weapons (though they do caution that secrecy in this area makes it impossible to draw “even remotely specific” conclusions). In particular, they argue that while certain components of a state’s nuclear C3I system, including early-warning satellites, may be vulnerable to cyber interference, the systems for communicating directly with strategic nuclear forces “are isolated and highly protected” and “in all probability, not vulnerable to cyber attacks.” If the absence of Russian writing on this subject is anything to go by, the Russian analytical community shares this perspective. Of course, in an effort to assess escalation risks, the most critical views are those of the Russian government and military, and it is unclear whether they are consonant with the perspective of nongovernment analysts.

In addressing these escalation risks, unilateral actions are the most realistic option, at least in the first instance. Some such actions, including revised war planning, would inevitably have to be kept secret, making progress difficult to gauge from the outside, but could prove quite effective nonetheless. After all, the escalation risks resulting from entanglement depend critically on which weapon systems a state procures, and on how these capabilities are deployed in peacetime and employed in wartime, as well as on political and military leaders’ understanding of how their actions are likely to be perceived by an adversary and their recognition of the challenges of correctly interpreting an adversary’s actions. Raising awareness of inadvertent escalation risks among the individuals responsible for strategic-level decisionmaking in a crisis, and factoring these risks into acquisition policy and war planning could, therefore, be a powerful approach to risk mitigation. To be clear, “factoring” escalation risks into policy and planning does not mean that they should always trump warfighting considerations. It simply means that they should be weighed up as part of the process of deciding whether a new weapon system or operational concept is in a state’s interest.

Ideally, China, Russia, and the United States would all embark on this process, and each should do so irrespective of whether the others do. Of course, given Chinese and Russian views that inadvertent escalation is unlikely, it is highly questionable whether Beijing or Moscow will do so—and there is only slightly more room for optimism that the current U.S. administration will devote much time or attention to this problem. However, all three governments should realize that they have nothing at all to lose from unilateral and secretive processes, and potentially much to gain. Nongovernmental analysts in all three countries could play a role by engaging with their governments—publicly or privately, as appropriate—to highlight the potential severity of the escalation risks.

A second step, which could run concurrently with these internal processes, would be intergovernmental discussions. Initially, the main purpose of a dialogue might simply be to assess escalation risks more accurately by better understanding a potential adversary’s perspectives. Given how important context is in shaping escalation risks, it would make sense to have separate U.S.-China and U.S.-Russia dialogues rather than a single trilateral process.

Reciprocity—each party feeling that a dialogue has actually enhanced its understanding of the other—will be key to sustainability. In this context, Zhao and Li break new ground by identifying specific areas on which China could explain its thinking to the United States. They argue that if Washington conveys “to Chinese leaders its recognition that mutual vulnerability is a fact, and that the United States will plan and posture its strategic forces on that basis,” Beijing should be willing to “explain its thinking on whether future hypersonic weapons would be armed with conventional or nuclear warheads, and on what space-based assets might be considered legitimate targets and under what conditions.” Arbatov, Dvorkin, and Topychkanov’s work suggests that these two issues—advanced conventional weapons and the survivability of space-based nuclear C3I assets—could also be fruitfully discussed in a U.S.-Russia dialogue. Interactions between cyber weapons and nuclear C3I systems could be a third potential focus for both U.S.-China and U.S.-Russia discussions. In the case of Russia, discussions might need to start with the very basic question of whether Moscow perceives there to be a “there there” at all.

The United States and Russia have agreed to discussions on strategic stability, and it is possible that the escalation consequences of entanglement could be discussed during these talks. Although the first round took place in September 2017, it is difficult to be optimistic that a substantive and productive dialogue will result. U.S.-China talks are likely to prove even more challenging to initiate for all the reasons explained by Li and Zhao. In the interim, track 2 discussions involving non-official participants could help fill the gap and hopefully pave the way for intergovernmental dialogues.

Over the long term, cooperative confidence building and even formal arms control could play an important role in risk mitigation—though their prospects are currently bleak (indeed, a prerequisite to such measures is intergovernmental discussions, and even they seem like a bridge too far right now). Nonetheless, governments can and should start their homework to develop and assess proposals.

For the United States and Russia, a good starting point would be three concrete proposals suggested by Arbatov, Dvorkin, and Topychkanov: transparency agreements that would preclude the “tacit massing” of platforms for delivering air- and sea-launched cruise missiles within range of the other’s “strategic targets”; an agreement to prohibit the testing and deployment of dedicated anti-satellite weapons; and the inclusion of intercontinental boost-glide systems under the central limits of a successor to the New Strategic Arms Reduction Treaty. Washington and Moscow should assess whether each of these proposals would be acceptable and, if not, whether it could be modified to make it acceptable. For example, Washington has long expressed concerns that a ban on anti-satellite weapons would be unverifiable. But, more focused confidence-building measures designed to protect satellites in geostationary or highly elliptical orbits (where key space-based nuclear C3I assets are located) may be more achievable. Washington and Moscow should also ask themselves which confidence-building concepts were sufficiently mutually beneficial to be negotiated on a stand-alone basis, and which would only make sense as part of a package that was balanced overall. Again, nongovernmental analysts could start to address these questions if governments do not.

Some of the challenges to taking these steps are all too evident; others are less so. Zhao and Li identify one challenge that has been frequently overlooked but is potentially serious: the question of blame. They note how many Chinese experts argue that it is the United States that is responsible for generating escalation risks and, therefore, that it is Washington that should take corrective measures. In a similar vein, some U.S. officials argue that because it was China’s choice to use some C3I capabilities to support both nuclear and non-nuclear operations, it is up to China to manage the consequences of this decision. It is all too easy to imagine a similar blame game being played by U.S. and Russian officials.

In the nuclear era, however, the concept of blame is moot. Given that the risks of inadvertent escalation are shared, so too should be the responsibility for managing them. It took the extraordinary dangers of the Cuban Missile Crisis, in 1962, for the United States and the Soviet Union to reach this realization and commence even a stop-start process of risk reduction. Unfortunately, leaders in Washington and Moscow seem to have forgotten this lesson. Meanwhile, given China has only very limited experience with serious nuclear crises, leaders in Beijing may never have learned it. Whatever real and serious U.S.-Russian and U.S.-Chinese disagreements exist, none of these states should want to reach the brink of a nuclear war—or go beyond it—before seeing the value of efforts to mitigate the risks of inadvertent escalation.


1 For this and other reasons, the escalation risks resulting from unmanned underwater vehicles seem—intuitively at least—as if they would be substantially more serious in a U.S.-China conflict than a U.S.-Russia conflict. Russia’s SSBN force is larger than China’s and spread over two oceans. Russia’s SSBNs appear to be substantially quieter than China’s. And, the naval chokepoints in the West Pacific are generally more restrictive than in the North Atlantic Ocean. However, more research is warranted to determine whether this intuition is correct. In particular, there is a very narrow waterway between the White Sea (where Russia’s Northern Fleet is based) and the Barents Sea, and the impact of potential U.S. UUV operations there deserves particular attention.