8. Conclusions and Recommendations

0 and 0

This study of Sino-American strategic relations finds that the United States and China each have the ability to cause the other grave harm not only by nuclear attack but also, and much more likely, by attacks on satellites and computer networks. While nuclear vulnerability is familiar, vulnerabilities in space and cyberspace are mounting as both countries increase their dependence on these domains for their prosperity and security. Technology is creating options for "nonviolent" and relatively low-cost strategic warfare, which could reduce inhibitions against attacks despite dangers of catastrophic results. Strategic defenses offer diminishing returns against large and advanced offensive capabilities, like those of the United States and China. Yet negotiated arms control of such capabilities is unpromising if not infeasible.

The United States cannot deny China a nuclear deterrent, and neither country can defend its satellites or networks well enough to prevent extensive economic damage if attacked by the other. Unlike nuclear weapons, attacking satellites and computer networks that support military operations could be of interest to both Chinese and U.S. militaries. Yet because neither space nor cyberspace is well separated into military and civilian sectors, escalation from tactical to strategic war in these two domains is a serious danger.

Conditions of mutual strategic deterrence, based on the futility of defense and credible threat of retaliation, either exist or are forming in all three domains, mainly because of technological and economic trends. In distinction from this, mutual strategic restraint signifies that the highest authorities of both states accept the imperative of mitigating national vulnerabilities cooperatively. It requires but improves on mutual deterrence: affirming Sino-American agreement not to initiate warfare in these domains and institutionalizing such agreement with CBMs, regular highlevel dialogue, and continuous contact through agreed channels to avoid miscalculation. By easing U.S. and Chinese concerns about the harm they might do to one another, mutual restraint can reduce fear, antagonism, and distrust in the larger relationship, further lowering the risks of conflict and strategic escalation.

In this spirit, the United States should propose an integrated approach to mutual restraint covering all three domains. While it can accept Chinese nuclear deterrence and bilateral nuclear no first use, it should do so contingent on Chinese agreement to extend the principle of mutual restraint to space and cyberspace. The framework the United States should adopt is summarized in table 8–1.

The Chinese may not yet appreciate how greatly China’s vulnerability will grow as its economy, integration, and power do. The PLA may also believe that attacking the satellites and computer networks on which U.S. forces depend is the only way to avoid defeat should war with the United States occur. Consequently, the Chinese might balk at and cyberspace, preferring agreed restraint only in first use of nuclear weapons. While this study develops the idea of mutual strategic restraint from an American vantage point, it concludes that China would be short-sighted to reject limits on attacking satellites and critical computer networks. Chinese leaders should realize that the United States possesses sufficient retaliatory capabilities to deter Chinese attack in these domains (a concept they know well from the nuclear field). Perhaps they will see enough strategic and political merit in broad-based restraint to surprise us by agreeing early on to explore it in earnest. In any case, it is in the U.S. interest to lay out its framework and pursue it with patience and persistence.

U.S. allies in East Asia could be ambivalent about Sino-U.S. attempts to limit their strategic vulnerabilities. On the one hand, they do not want Sino-U.S. tension, arms races, or hostilities at the strategic level, because they too have vulnerabilities. On the other hand, they would not want to relieve Chinese fears that using military force could escalate into these strategic hostilities. The United States can allay regional concerns about such strategic decoupling by renewing its security commitments, maintaining its presence, and insisting that Sino-American strategic restraint also apply to allies. U.S. extended nuclear deterrence of Chinese nuclear threats to U.S. allies would be unaffected but nevertheless should be reaffirmed unequivocally. In addition, by establishing that mutual restraint in all domains covers allies, the United States would in effect be extending deterrence in space and cyberspace as well, in the sense that China could face U.S. retaliation.

The United States should be able to deter China from attacking Taiwan by conventional military escalatory options other than space and cyber escalation, much less nuclear weapons. U.S. escalation to strategic warfare would have dire consequences, including Chinese retaliation, and is not the right solution to the problem of improving Chinese conventional capabilities. In general, the United States should state its expectation that mutual strategic restraint should strengthen overall stability and security in the region—in effect, putting down a marker that Chinese conventional aggression in the region would prompt the United States to rethink mutual strategic restraint.

A Future without Sino-U.S. Mutual Strategic Restraint

This is an ambitious proposal for a Sino-American relationship that is already loaded with weighty issues, and for a region that is already unnerved by China’s growing power. Moreover, though it is possible that Chinese leaders would see the virtues of mitigating national vulnerabilities by mutual restraint, it is as likely that they will be hesitant because of the sheer scope and significance of the framework and the PLA’s aversion to foreclosing military options in space and cyberspace. Therefore, the U.S. administration that offers this framework may not be the one that sees it bear fruit. Moreover, current political and economic conditions in the United States are not auspicious for a major initiative premised on the idea that China can be trusted. If there is a reason for the U.S. Government to take this step, it is strategic, not political. Therefore, it is only fair for U.S. policymakers to ask why it is important to propose this now and what harm would come if they do not.

The failure to pursue and achieve some form of agreement on mutual strategic restraint could have several adverse results in the years to come, ranging from the unlikely but traumatic to the likely but subtle. The first is that China could unleash major attacks on U.S. satellites or computer networks in the context of or as a prelude to a confrontation over, say, Taiwan or Korea—perhaps in hopes of deterring the United States from armed intervention or out of fear that the United States might launch such strikes on China preemptively. Instead of giving the United States pause, such attacks would likely trigger U.S. retaliation, leading to a spiral of attacks and counterattacks, sending strong shock waves through the Chinese, U.S., East Asian, and world economies.

Even in the absence of armed conflict, China might mount a disruptive attack—beyond intelligence collection—on one or more critical U.S. computer networks, perhaps in response to some U.S. action perceived as highly provocative (such as the sale of advanced weaponry to Taiwan or support for Tibetan independence). In that event, the United States might well retaliate in kind, since failing to do so would leave deterrence in ruins and expose the United States to further attacks. While not necessarily leading to general war in cyberspace, this scenario suggests that both countries could suffer large-scale economic and societal damage unless mutual restraint was agreed and observed.

Short of actual hostilities in space and cyberspace, the absence of agreed mutual restraint and the corresponding growth in both countries’ vulnerabilities in these domains would erode U.S., as well as Chinese, security. Both countries might feel compelled to invest more in both strategic offensive and strategic defensive capabilities, even though the former would increase vulnerabilities while the latter would fail to mitigate them. Absent mutual restraint, neither country would feel it could afford to allow the other to gain advantages.

Faced with unchecked Chinese strategic threats, it is not difficult to imagine the United States spending a growing portion of its defense budget on missile defense, satellite defense or redundancy, and computer network defense or redundancy. To illustrate, were the United States to double spending on missile defense (now about $10 billion annually), double the number of working satellites to achieve space security through redundancy, and triple the budget of U.S. Cyber Command (at least $3 billion per year), it could add about $200 billion to the defense budget over the next decade.¹ As this is written, the U.S. Government is being forced by its fiscal crisis to consider options to reduce defense spending by $400 billion to $800 billion in the coming decade. Adding new requirements for strategic capabilities would preclude such reductions or require cutting into the muscle and readiness of U.S. conventional military forces. Even if increased U.S. investments to address strategic vulnerabilities were lower than this illustration, fiscal conditions could hardly be worse for any such expenditures.

Given offense dominance, a U.S. commitment to "sufficient" strategic defense would be inherently open ended and potentially self defeating: it would induce China to invest more in strategic offensive capabilities to frustrate improvements in U.S. defenses. China faces a similar calculus: while it is increasing defense spending at double-digit annual rates, it could be pressured to commit more of these resources to competition with the United States in both strategic offensive and defensive capabilities.

Meanwhile, the increase in U.S. national vulnerabilities would accelerate. The growing precariousness of assured access to space and cyberspace could sap confidence in these domains and in the economic opportunities and benefits they afford. Awareness of the potential for large-scale disruption would weigh at least psychologically on markets and on the U.S., Chinese, and world economies. In parallel, as each country becomes more reliant on space and cyberspace for national security, both would experience sagging confidence in the ability of their forces to operate. Although each power might take some comfort in the other’s military doubts, U.S. use of space and cyberspace for national security is becoming too vital to trade it for denying Chinese use of those domains.

Of course, it is possible, even probable, that conditions of mutual deterrence will emerge in all three offense-dominant domains even in the absence of cooperative, institutionalized restraint, making such strategic warfare less likely for both countries. This begs the question of what harm could come from having mutual deterrence without an agreed framework of mutual restraint recommended by this study. One danger is that unless both sides acknowledge deterrence and accept restraint, the risks of miscalculation and breakdown could be high at moments of stress. Another is that the United States could waste resources on strategic defenses in the absence of Chinese agreement not to attack first and the concrete CBMs and attendant notifications and consultations that could buttress such agreement.

Relying exclusively on fear-based deterrence without agreed cooperative restraint could also affect Sino-U.S. relations. This study began with the observation that Sino-American interests are sufficiently divergent to make mutual restraint important but also sufficiently compatible to make mutual restraint possible. The Soviet-American model was of two powers whose threats and fears of destruction caused them to set limits on their otherwise antagonistic relationship—threats and fears that occupied the core of their relationship, and limits that did little to reduce antagonism. Mutual deterrence permitted the United States and Soviet Union to carry on their struggle with less danger of it getting out of hand. For the United States and China, mutual strategic restraint is a way to replace or at least assuage fear with trust and to create more space at the relationship’s core for cooperation.

Mutual strategic restraint does not guarantee that Sino-American relations will be free of friction and deliver results that invariably serve U.S. interests. However, if the United States does not offer a framework for agreed and broad-based restraint, especially as Chinese attitudes about these issues have yet to solidify, it will lose an opportunity not only to mitigate vulnerabilities but also to set conditions for the sort of relationship it claims to want. Though these concerns may not seem urgent, it is better to commence the pursuit of mutual restraint than to wait for national vulnerabilities to grow, to hope that Chinese views will not toughen, and to count on the illusion of affordable strategic defense.

In essence, mutual restraint is about the ability of humankind to foresee and manage the effects of its inventions. The advent of nuclear weapons last century was accompanied by warnings—from none other than their inventors—that that technology would require unprecedented international openness and creativity to avert destruction.² Humankind’s discoveries and use of space and cyberspace are defining features of this century’s global economy and society. Whether their promises and potential will be fulfilled or instead turned to destructive ends depends above all on whether the world’s leading power and its fastest rising power can find common ground and lead. Although these proposals do not require supranational authority, as early proposals to contain nuclear risks did, they do require these two powers to accept limits on their strategic freedom of action.

The Longer Term

Although the timeframe of this study is the coming decade or so, its prescriptions could apply beyond that, assuming that offense dominance will persist and that the United States and China remain world powers, vitally dependent on and vulnerable in space and cyberspace, with both convergent and divergent interests. Take away either of these basic conditions, and the logic of mutual strategic restraint will fracture. On the other hand, Sino-American strategic restraint can perpetuate the conditions that produce this logic. By mitigating the dangers of strategic conflict, mutual restraint should help the United States and China fully exploit space and cyberspace despite their vulnerability. Moreover, it can free up the two countries to develop a more cooperative relationship. The framework offered here, if actively managed by the United States and China, can contribute to their mutual prosperity and security even as conditions evolve.

Still, given accelerating changes in technology, predictions beyond this study’s timeframe of a decade or so are more art than science. Although current offense dominance in strategic domains results from discernable trends in information sensing, processing, and sharing (thereby enabling effective targeting), it is possible, if improbable, that these technologies will come to favor defense more than they have to date. Offense is not destined by laws of either physics or economics to prevail.

For example, missile defense could become easier and cheaper even against large and sophisticated attacks. Space-based sensors, directed energy, and especially breakthroughs in data processing (for example, quantum computing) could vastly improve target tracking and interception, even against large and complex attacks. If it is relatively easy to intercept a satellite with a predictable orbit, it could also become easier eventually to intercept an object in an unexpected trajectory with little warning. Still, the prospect that a single warhead would penetrate and do unacceptable damage will perpetuate offense dominance in nuclear warfare despite improvements in defense technology.

Offense dominance in space could be eroded by the placement of numerous cheap decoys, though they would have to have signatures resembling real satellites and would likely be revealed as decoys given enough time to observe them. Maneuverable or stealthy satellites would obviously increase the difficulty of targeting, at some cost in dollars and performance. Resiliency could be gained by distributing missions and functions among a large number of satellites instead of concentrating them among small numbers of high-value/high-cost platforms, as has been the general practice. Meanwhile, it can be assumed that hardand soft-kill technologies will also progress, leaving access to space a serious vulnerability.

Information technologies are especially unstable and unpredictable, which raises questions about the persistence of offense dominance in cyberspace. Just as useful computer networks tend to be open and accessible, they can also become more self-aware, sensitive to intrusion, adaptable, and thus resilient. Even large and sophisticated attackers could find it increasingly hard to cause widespread and lasting disruption. Moreover, the concept of "dynamic defense" implies that attackers can be detected and neutralized the instant they gain unauthorized network access. Defenses could become sufficiently cost effective that even advanced attackers could be deterred not only by the fear of retaliation but also by the prospect of failure.

Such speculation about a shift toward defense dominance runs against what may be enduring features of the digital age: steadily increasing abilities to pinpoint objects on and near the Earth, to share that information, and to guide other objects to those same points—in a word, targeting. Current inventive and price performance trends in sensing, global positioning, processing, and transmission technologies suggest that targeting will, if anything, improve. By the same token, exceedingly strong market forces that favor access and collaboration will likely confront network defenses with a growing challenge.

While this study has found that neither the United States nor China can buy its way out of vulnerability through investment in strategic defense, it goes without saying that both will at least continue exploratory, competitive research for ways to overcome offense dominance. At the same time, if Sino-U.S. mutual strategic deterrence works, if mutual restraint is institutionalized in both countries and between them, and if trust grows, the urge to "break out" could decline, even if science makes it more possible. Moreover, while deterrence is static and potentially fragile under technological stress, mutual restraint can be adapted cooperatively to preserve stability despite technological stress. While technology more than a decade out is unpredictable, the safe bet—and a prudent assumption—is that the world’s leading states will still be grappling with the vulnerabilities produced by their own power and global integration.

Recommendations

Beyond the core recommendation for the United States to offer a framework of concepts and terms for integrated mutual restraint, several concrete suggestions for U.S. policymakers come to mind. Initial diplomacy on actual measures of restraint would not be with China but with allies, who must be given the opportunity to contemplate and comment on U.S. views of vulnerabilities in these strategic domains, of their significance for the region, and of negotiating with China. In this regard, it is important to impress upon allies that strategic restraint is part of a larger U.S. strategy to maintain stability and eliminate the danger of destructive war in this vital region.

It would also be good early on to share with Chinese counterparts U.S. analysis of vulnerabilities in space and cyberspace, especially between large and sophisticated potential adversaries. It is not clear that Chinese leaders understand how China’s growing and irreversible reliance on satellites and computer networking exposes it to possible economic harm. Dialogue could help educate them about this reality, while stressing that the United States also faces vulnerabilities and is not seeking advantage at China’s expense. The United States should share with Chinese political and economic elites an alternative to the PLA view of strategic reality, including the risks and consequences of escalation in space and cyberspace. Unofficial dialogues on space and cyber security might be a good means of engaging a broad range of Chinese civilian and military actors and making them more aware of how their interests might be affected. It would also be timely to tell the Chinese that the United States is willing to consider a bilateral nuclear no-first-use pledge provided China is willing concurrently to discuss similar ideas about space and strategic cyberspace.

Regardless of progress on terms of mutual restraint, it is important that operational decisions that could lead to hostilities in any strategic domain have strong political oversight. In regard to the use of ASAT and offensive cyber capabilities, the United States should review its protocols for such oversight and delegation of authority to military commands under peacetime and wartime conditions and should urge Chinese civilian leaders to do likewise. Finally, the need for the United States to speak with one voice on these matters argues for intense civilian-military, executive-congressional, and bipartisan discussions.

Such steps should be taken not with undue urgency but with care, composure, and conviction that a regime of mutual strategic restraint is right for the United States, for the security of a vital region, and for putting Sino-American relations on a stable long-term strategic footing. At this formative stage in what will be the world’s most important relationship for generations to come, the United States cannot afford to be passive and reactive. Because the relationship is so complex and fluid, and because vulnerabilities are growing, a better time may not come for an American initiative to offer a framework for strategic stability.

In the backdrop of such deliberations and diplomacy, the United States should be increasingly vocal and clear on the matter of strategic deterrence. Just in the past year, the U.S. Government has indicated that it will strengthen deterrence against attacks on satellites and on computer networks. This is a two-edged message: the United States will not be the first to launch strategic attacks in these domains, yet it can and will retaliate if others strike first. Even as the United States becomes comfortable with the ability of China to deter U.S. nuclear attack, it should make the Chinese uncomfortable about the consequences of satellite or cyber attacks.

This study is not meant to be the final word on the paradox of power and the need for mutual strategic restraint. We urge more debate on the strategic concepts illuminated here, not only in the United States and in China but also between Americans and Chinese. In addition, a number of questions merit further examination:

What missile defense capabilities would afford assured protection against small, hostile nuclear weapons states or unauthorized missile launches without raising doubts about the acceptance of mutual deterrence between the United States and China?

How can computer networks used for military C⁴ISR be partitioned from those that enable civilian and commercial information-sharing so that escalation firebreaks can prevent unwanted and potentially cataclysmic general war in cyberspace in the event of crises or hostilities?

What concrete CBMs beyond those proposed here could buttress trust in Chinese and American acceptance of mutual restraint in the use of offensive capabilities?

What specific methods of Sino-American notification of third-party or ambiguous attacks in space and cyberspace could increase assurance against mistakes and miscalculation?

Under what conditions could other states with offensive strategic capabilities, such as Russia, subscribe to the principles and terms of mutual restraint?

Perhaps mutual strategic restraint will prove to be more of a process than a set of definitive obligations—a process in which political leaders and military commanders come to grips with the need to act prudently and to treat strategic vulnerability as a common problem to be solved cooperatively. One way or another, as technology and integration expose even the most powerful nations to growing threats from one another, the United States and China have strong interests to partner and special responsibilities to lead. We hope this book will prompt them to do so.

Notes

1. The United States has spent on average $10 billion per year for the past 10 years on total missile defense (including research and development, military construction, and procurement). Doubling this expenditure for the next 10 years would put the missile defense program at $200 billion. A total of 424 satellites, 208 working, have been deployed with a total expenditure of $62.3 billion. At a low end cost of $147 million per launch, deploying an additional 200 satellites for redundancy would cost $30 billion. In the cyber domain, as technology gets better, expenses to defend U.S. critical networks will increase. U.S. Cyber Command, not fully operational until October 2010, requested $3.2 billion for its 2011 budget. Lacking personnel and facilities, this budget is likely to increase over the next 10 years to parallel increased cyber warfare capabilities. If the budget is tripled, expenditures would rise to $96 billion over a 10-year period, rivaling current costs of missile defense. Budget figures were derived from the Congressional Budget Office and Congressional Research Service.
2. For instance, Robert J. Oppenheimer’s insistence that U.S. use of atomic weapons against Japan be followed by a proposal to internationalize the knowledge control of nuclear power.